118 | Adam Riess on the Expansion of the Universe and a Crisis in Cosmology

John D. & Catherine T. MacArthur Foundation

Astronomers rocked the cosmological world with the 1998 discovery that the universe is accelerating. Well-deserved Nobel Prizes were awarded to Saul Perlmutter, Brian Schmidt, and today’s guest Adam Riess. Adam has continued to push forward on investigating the structure and evolution of the universe. He’s been a leader in emphasizing a curious disagreement that threatens to grow into a crisis: incompatible values of the Hubble constant (expansion rate of the universe) obtained from the cosmic microwave background vs. direct measurements. We talk about where this “Hubble tension” comes from, and what it might mean for the universe.

Support Mindscape on Patreon.

Adam Riess received his Ph.D. in astronomy from Harvard University. He is currently Bloomberg Distinguished Professor and Thomas J. Barber Professor of Physics and Astronomy at Johns Hopkins University and a Senior member of the Science Staff at the Space Telescope Science Institute. Among his many awards are the Helen B. Warner Prize of the American Astronomical Society, the Sackler Prize, the Shaw Prize, the Gruber Cosmology Prize, the MacArthur Fellowship, the Breakthrough Prize in Fundamental Physics, and the Nobel Prize.

[accordion clicktoclose=”true”][accordion-item tag=”p” state=closed title=”Click to Show Episode Transcript”]Click above to close.

0:00:00 Sean Carroll: Hello, everyone. Welcome to the Mindscape Podcast. I’m your host, Sean Carroll, and today we’re going to do the rare thing here on Mindscape and actually talk about something that I’m supposed to be an expert in. And we talked to Netta Engelhardt a little while ago about black hole information, and I’ve actually written papers about black hole information, but it’s never been the central concern of my research. It’s always been sort of a side light for me. One thing that I have been very, very concerned about is large-scale cosmology, the expansion and acceleration of the universe.

0:00:29 SC: So today we’re extremely happy to have as our guest Adam Riess, who was one of the discovers of the acceleration of the universe, and who has been noting that there is another cosmological anomaly, known as a Hubble Tension, that calls into question the beautiful fit that we have in our cosmological models that we’ve been very proud of over the last 20 years since the acceleration of the universe was discovered. As Adam alludes to in the podcast, he and I were graduate students together in the Astronomy Department at Harvard at the same time. Adam was a floor below me in Perkins Hall. My office mate was Brian Schmidt, who went on to lead the High-Redshift Supernova Team. So Adam and Brian were two people who shared the Nobel Prize with Saul Perlmutter for discovering the acceleration of the universe.

0:01:13 SC: So I like to think that even though I had nothing to do with the actual research, that I at least put them in the mindset of thinking about the acceleration of the universe. I was interested in it at the time. I wrote a review article with Bill Press and Ed Turner about the possibility that there could be a cosmological constant, though to be perfectly honest, I was absolutely not saying in the early 1990s that there probably is a cosmological constant or dark energy or the acceleration of the universe. You would have won money from me if you had tried to bet me at even money and you had taken the side that there was a non-zero cosmological constant. I was skeptical, and in my review article, I gave you all the reasons to be skeptical. I said, “Here’s why it would be really, really weird if this were actually true.” Adam and Brian and Saul and their two big and very, very talented teams showed that there was.

0:02:03 SC: So that was a huge surprise. And as we talk about in this podcast, it was the kind of thing, even though it was a huge surprise in 1998 when they discovered the acceleration of the universe, we instantly had a good explanation for the surprise. The idea of a cosmological constant had been proposed by Einstein in 1917, and this is why in 1992, we could write a review article about it. Even though we were skeptical that it was there, we knew what it would be like if it were there. And when the High-Redshift Supernova Team and the Supernova Cosmology Project got evidence that the universe was accelerating, we could instantly go, “Ah, here’s the explanation. Let’s think about this.” And so in some sense, the observations were confirmed by theory, as they say.

0:02:45 SC: Nowadays Adam is pushing hard on a different kind of discrepancy between expectation and reality. There’s two different ways of thinking about measuring the expansion rate of the universe, so not the cosmological constant, but the Hubble Constant, the current rate at which the universe is expanding. You can just do it directly, like Hubble did, looking out into the sky and measuring the recession velocity of galaxies, or you can infer it from bigger picture measurements of cosmology, including the Cosmic Microwave Background. It turns out, as Adam has been emphasizing for a couple of years now, these two methods give different answers. And closer inspection has not brought them into closer alignment, it has driven them apart. It has made the discrepancy even larger.

0:03:29 SC: The problem now is that this is not an experimental result that is confirmed by a theory. We don’t have any real good obvious theoretical ways of dealing with this. So Adam is also a great communicator. We’ll do a very fun job of building up slowly the basics of cosmology to what this discrepancy really is about and some of the ways you could possibly try to cure it. We don’t know if it’s real. We don’t know what the cure would be if it were real, so this is very much an ongoing research problem in cosmology.

0:04:01 SC: Let me remind you if you don’t already know that we have a Patreon for supporters of the Mindscape Podcast. You can go to patreon.com/seanmcarroll, and what you get for throwing a few bucks my way is you get monthly Ask Me Anything episodes. So we’re trying to make the Ask Me Anything episodes publicly available to listen to, but if you actually want to ask questions, you have to support on Patreon. That’s where the question-asking happens, and the other thing is you get ad-free versions of the podcast. I don’t think the ads are very objectionable even if they’re there, but if you like your podcast experience to be completely pristine and uninterrupted, you can get that at the Patreon site. But again, as I always like to say, if you don’t want to support, that’s completely fine. It’s completely optional. We love you anyway. Thanks for listening to the Mindscape Podcast. That’s the most important thing. So with that, let’s go.

[music]

0:05:07 SC: Adam Riess, welcome to the Mindscape Podcast.

0:05:09 Adam Riess: Thank you for having me.

0:05:10 SC: So it’s a big universe out there. I know that you’ve been charged with figuring it, how big it is, how fast it’s expanding, things like that. I think, I’m hoping that over the course of this podcast, we can sort of go from the very basics to some pretty deep stuff, but let’s start with the basics. How do you think about modern cosmology? What does the person on the street need to know about the universe on the larger scales?

0:05:32 AR: Right, well, they should know that the universe is expanding, that it looks like we live in a fairly ordinary place in the universe. And one of the most, I think, profound and ambitious endeavors we take as humans is to understand the universe, to make measurements of it and understand how old it is, what its fate is, what it’s composed of, and we do these things both observationally and theoretically.

0:06:00 SC: On the one hand, we say that it’s smooth, it looks more or less the same everywhere, but that’s only on very large scales. It’s lumpy with galaxies and stars on smaller scales.

0:06:08 AR: Correct. It’s just like looking at the Earth. Any particular spot, it might be a lake, it might be a mountain, but if you take a big enough chunk, then you get a bunch of lakes, a bunch of mountains, the various things, it starts to average out.

0:06:21 SC: And so if we… Why don’t you just… Again, for the people who really… I don’t know if I have any listeners who fit into this category, but for the people who really don’t know, if the universe is expanding now, it must have been smaller in the past. What do we know about that?

0:06:35 AR: Yeah, well, the evidence to us that that is the case is when we look out, we look back in time, so we can literally see those times when the universe was more compact and we see it was hotter, as you expect. When things are close together, they get hot, and when they separate, they get cool. We also see things generally age in the universe, which tells us that it really was younger in the past, and it is older now. And of course, when we look at all the objects around us, we see their light has been stretched or redshifted by the expansion of space, and so if those motions of the objects around us were just random, like you get with the Doppler effect when something is moving toward or away from you, it could go either way. The light could get blueshifted or redshifted, but we see pretty much everything is redshifted, which tells us that it must be all of space that is expanding.

0:07:30 SC: I do, even at this late date, sometimes get questions about, “Are we sure the universe is expanding?” versus what you and I would call a tired light kind of perspective, like the light just loses energy into the universe as time goes on. Do you have a favorite response to those kinds of questions?

0:07:48 AR: Yeah, one of my favorite observational consequences of expansion that wouldn’t happen if you just had tired light is the fact that distant objects, not only do they move away from us quickly, but the events that occur far away are stretched or dilated due to the expansion of space, so that’s another consequence that wouldn’t happen with tired light. And indeed, when we look at, for example, a certain kind of supernova explosion, a Type Ia supernova, we know what the evolution of that explosion is very well. It takes 20 days to go from explosion to maximum, and we see that period elongated by exactly the amount you would expect because of the expansion of space, in some cases, up to double or even triple the normal time interval. So we have some very good evidence that the universe is expanding.

0:08:39 SC: And actually, that’s a good example of I think a paradigm that’s going to come back again and again in the conversation, which is that you have an idea like the universe is expanding and you don’t just test it in one way. These are systems that fit together in many, many different ways, and the whole paradigm of the expanding universe according to general relativity could have failed in many, many different ways, and we’ve looked at it and so far it’s on the right track.

0:09:04 AR: Yes, that is absolutely true.

0:09:07 SC: The other thing that I think that people don’t appreciate maybe as much as I would like them to is it’s kind of easy, relatively speaking, to measure the redshift to a galaxy or something like that, whereas it’s really hard to measure the distance, and this is basically your stock in trade now, right, measuring distances to faraway objects.

0:09:26 AR: Yeah, and it’s very non-intuitive. If you look at a faraway car, and I ask you, “How far away is that car?” you have a good sense of how big a car is. And so the fact that’s it’s an itty-bitty car, you say, “It looks like 100 yards away.” If I ask you, “How fast is it traveling?” you would have a very hard time judging that, especially if it’s coming directly at you. And so what is easy for us here on Earth and what is hard is exactly sort of inverted, I would say, when we look out in space. And primarily that’s because the objects we see in space, unlike cars, are not familiar to us. We don’t get to walk right up to one and sort of eyeball it and say, “Oh, so that’s how big a car is.” In our case, it might be a galaxy, it might be an exploding star, but we don’t get to walk up to these things and figure out intrinsically their properties, how luminous or how big, and so we always struggle a little bit with figuring out when we see an itty-bitty one or a big one one far away just how far away that is.

0:10:29 SC: Right, because cars are either more or less the same size or at least we think we know how big they are, whereas the universe is full of messy, non-standard things like stars, like you just said. Stars could be big, stars could be small. So what do we do about this? I guess the idea is if we have some standard object that we know how big it is intrinsically or we know how bright it is, then by measuring how big it appears or how bright it appears, we can figure out ex post facto how far away it is.

0:10:56 AR: Right. That’s right. And so that is our challenge. Of course, we have tools. One of our favorite tools really goes back to the Greeks and is the idea of parallax. As the Earth goes around the sun our perspective on relatively nearby objects changes with respect to more distant objects, same experience we all have if you take a walk and look at how nearby trees move a lot and faraway trees move very little, and that is just basic trigonometry. All you have to go is know the diameter of the Earth’s orbit, which we know, and measure those small angle and you can figure out exactly how far away the nearest stars are.

0:11:41 SC: And is that generally what you do or that’s just the first step on the ladder, right?

0:11:45 AR: That is where we start in many cases. We measure how far away the nearest stars are. Unfortunately, it turns out stars are very far away, and so the angle, the parallax angle, that angle through which a star moves, often becomes very, very small. It becomes a milliarcsecond, for instance, might be a typical value for the kind of stars we’re interested in. And so therefore, we can really only measure the stars sort of nearby to us, how far away they are, and from there we sort of have to bootstrap, we have to find, if we’re lucky, a very luminous version of one of those stars, measure its distance, infer its luminosity, and then if we’re lucky, we can use a powerful telescope like Hubble and see one of those exact same kind of stars in another galaxy, perhaps a galaxy that hosted a supernova explosion. And so that allows us to calibrate the luminosity of the supernova which those we can see much further away.

0:12:47 SC: So I guess what you’re sort of hinting at, but maybe it’s worth laying on the table, is the idea of this distance ladder that we can see the parallax for the stars nearby, but for the galaxies that we care about receding from us and of cosmological interest, there’s essentially zero chance we will ever measure the parallax there, so we have to bootstrap our way up.

0:13:08 AR: That’s right. That’s right. And so this concept of a distance ladder, which has been around a long time in cosmology, is really… It’s not based on physics. It’s really just geometry, and this simple idea that if two things are co-located then if you understand what one of those objects is, if you can use that to gauge the distance, then you essentially calibrate the other object that it’s co-located with.

0:13:33 SC: It does, though… It begins to make us think that this is a tricky business, right? There’s a lot of steps along the way, and there’s going to be little errors at each step. So you gotta work very, very hard to make sure that this is… Every floor of the very tall building is firmly established.

0:13:52 AR: That’s correct. That’s correct. And so one of the trickiest parts is to make sure that when you build a distance ladder that the nearby version and the distant version of the same object you’re looking at, whether that be a star or a supernova, that those are really the same objects. And of course, as you put together a number of steps, any error you make in one step will propagate up the measurement chain. And so that’s why in the old days, when people built distance ladders, it was a rickety distance ladder, I would say, with 8, 10, 12 rungs. And now we’ve gotten it down to just about three rungs, and so that really simplifies things quite a bit. Another significant improvement is we’ve been able to make those measurements in a more homogeneous way. We use now the same telescope to observe, for example, the nearby stars and those distant stars in other galaxies, and so the problem of calibrating your telescope sort of drops out in the analysis.

0:14:56 SC: Yeah, I wouldn’t even have thought that that would be a problem, but yes, now that you mention it, it sounds like a big one. So, okay, you’ve mentioned supernovae. They’ve been very important to your career. Why don’t you tell us a little bit about how the supernova can be used as a distance object? I remember when I was in graduate school, the idea that we would even look for supernovae rather than just being lucky enough to see one was kind of heretical.

0:15:18 AR: Right. Well, we’re very fortunate that there are not many macroscopic objects in the universe that are truly standard, that are more or less the same. But there is a limiting mass for a certain kind of star, a very compact star when its material is in a degenerate state, known as the Chandrasekhar limit, named after, of course, Chandrasekhar who showed at least back in the 1930s that there was a limit about 1.4 times the mass of our sun. And if a star reaches that limit and surpasses it, we believe you get a runaway thermonuclear explosion. You have the conditions for fusion all over the star, and so this is kind of a standard bomb. So we believe that there are these stars that exist below the Chandrasekhar limit, but that they live in binary systems, that there are other stars next to them.

0:16:13 AR: And some time over the life of those two stars, mass may transfer from one star to the other. That could happen because one star swells up, it could happen because they merge, but either way, that white dwarf we call star crosses that Chandrasekhar limit and explodes, and that gives us a very uniform explosion. And so when we see these objects and recognize them far away, we can infer their distance from their brightness.

0:16:42 SC: Right. And I think mostly what I want to do here is talk about the Hubble Tension and the modern stuff, but while we have you here, I can’t help but ask a little bit about the famous discovery in 1998 of the accelerating universe. So this was something that was done with supernovae playing a big part. You were a part of the team, won the Nobel Prize, etcetera. Tell us number one, what you did as a scientist, but maybe also what was going through your head when you first realized, “Oh, my goodness, the universe is accelerating.” This was not something that I expected anyway.

0:17:14 AR: Well, actually, to be honest, you were part of it. I’ll mention that in a second. But so our team had been measuring nearby Type Ia supernovae and gauging how fast the universe was expanding by measuring their distances and redshifts. What was new by the mid-1990s, it became possible for the first time with bigger telescope and bigger detectors to find distant supernovae on demand. That is you search a wide swath the sky and return to that point a month later or so, digitally subtract images and find a new exploding star. But these were very faint ones, very distant ones that were telling us how fast the universe was expanding in the past, maybe 5-7 billion years ago. And by comparing the past expansion rate to the present expansion rate, we could infer how much the expansion was slowing down, at least that’s what we thought, that it would be slowing down, the attractive gravity of all the stuff in the universe would slow it down.

0:18:17 AR: So in 1998, I was a post-doctorate fellow working on the High-Z Supernova Team and was pulling together this data, and I wrote a simple computer program to tell me, okay, what is the mass of the universe that would cause this much deceleration, not yet realizing that that’s not what was going on. And so my little computer program I wrote, reported back that the best fitting mass to the universe was a negative mass. And of course, there’s no such thing, but now I like to say, “Computers don’t know physics. They just know whatever equation you give them.” And so after a few days, of course, realizing, “This is really un-physical. There’s no such thing as negative mass,” I pulled up a review article that you wrote in 1997, was it?

0:19:06 SC: I think it was ’92.

0:19:07 AR: Which year was it?

0:19:09 SC: ’92, I think.

0:19:10 AR: ’92, I’m sorry, 1992. A review article that you wrote talking about the cosmological constant, its long and tortured history of this repulsive gravity term that could cause the universe to expand faster and accelerate. And so I pored through that, found the relevant equations, [chuckle] and re-fit the data using those. And lo and behold, the data was arguing very strongly for the need for something like this cosmological constant in lieu of negative matter. And I always think that’s kind of a funny way to think of it because in the simple equations, they would almost look the same, except one of them is potentially physical and the other is not.

0:19:56 SC: So the best fit was a positive amount of matter, but this repulsive force due to the cosmological constant.

0:20:01 AR: Correct. About 30% in matter and about 70% in the form of this cosmological constant or dark energy.

0:20:11 SC: But again, it was a surprise. Did you have to sort of… We’ve all written computer programs that have gotten the wrong answer, [chuckle] so there must have been a lot of tearing out of hair.

0:20:20 AR: Yeah, that happens much more often when one makes a discovery, so I certainly spent a while trying to figure out what was going on, feeling, “Oh, obviously, I’ve made some kind of mistake.” But it’s quite frustrating when you look for a mistake and you don’t find it, and this is the value of being on a team with smart people is at some point you realize, “I’m just never going to find this bug, but somebody else will not make the same mistake or have the same bug.” So I passed the data to Brian Schmidt, who you know quite well.

0:20:50 SC: Yeah.

0:20:51 AR: He was your office-mate and the leader of our team, and he re-fit the data, writing his own program and he got the same answer. And then at some point, we started getting enough confidence in this, and the rest is history.

0:21:06 SC: But the result came from measuring these supernovae. Like you said, roughly speaking, you expect them to be uniform because the Chandrasekhar limit is uniform. That’s not exactly true, but you can hopefully correct for that. But how worried were you that there was just something weird going on with supernovae, not with the cosmological constant?

0:21:26 AR: Right. We studied that pretty carefully, so my thesis had been about measuring distances to supernovae and correcting for annoying effects like dust in galaxies that might block the light of supernovae and fool you into thinking they were further away than they really are. So we had used those techniques and related techniques to account for things like that. What we worried about was whether supernovae born when the universe was younger were just different than the ones born today. But fortunately, even nearby, we have analogs for what I would call old and young supernovae. We have galaxies that are essentially what we call old galaxies. They only have old stars. They’re like senior citizens’ homes. And then we have young galaxies, which primarily have young stars, and so those are spirals and ellipticals and nearby we have Type Ia supernovae in both those kinds of galaxies. And so we were able to demonstrate and convince ourselves that our techniques gave distances to those galaxies that were agnostic about what kind of galaxy and how old they were.

0:22:35 AR: And so it finally came down to saying to ourselves, “If the supernova doesn’t know how old it is in its galaxy, how does it know how old it is in the universe?” And so at some point, you do the best you can, and eventually you go ahead and you publish those results. And of course, what was really convincing to not just us but the community was that other techniques completely unrelated to supernovae reached the same conclusion within about, I think, two or three years.

0:23:06 SC: Yeah, which makes this a perfect segue. So by the way, congratulations. Accelerating universe, that’s pretty awesome. But you did not rest on your laurels. You’ve kept going, and it makes sense at this point to bring in a whole another part of the story because, like you said, it was verified by independent methods and one of the big methods was the cosmic microwave background. So why don’t you tell us that story a little bit?

0:23:27 AR: Sure, so the cosmic microwave background provides us a completely different, I would say, measurement chain, but also a way of looking at the universe. We can see the radiation, the heat left over from the Big Bang, and there are structures in that heat left over that allow us to calibrate in a completely different way the state of the universe. And so, I told you earlier that we had measured about 30% of the universe in the form of this matter and about 70% in the form of this dark energy, reaching this sort of 100%, if you will. And the cosmic microwave background basically came along and said, “Yeah, we see that 100% too.” They have different sensitivity, so they couldn’t initially break it into which part was matter and which part was energy. But astronomers had for decades measured the amount of matter in the universe to be that 30%. And so really the only options were that there was the 30%, and this 70% existed in this cosmological constant or it didn’t. And so the fact that the cosmic microwave background came in and measured and very precisely, I should say, measured the sum total to be the 100% within a couple of percent really cinched it at that point.

0:24:44 SC: Okay. But we’ve done our due diligence in explaining that the universe is expanding and stuff like that to the people who are not experts, so let’s roll up our sleeves and get in the details a little bit. How in the world does measuring light left over from the Big Bang tell me anything at all about what the matter or energy content is today?

0:25:04 AR: Right, well, you might think about it like this. It tells us very much about what the state of the universe was, how much matter and energy and how fast it was expanding shortly after the Big Bang, which is an awfully long time ago, more than 13 billion years ago. And what’s so great about physics is if you understand physics well and you measure the state of an object, maybe a moving object through your room, if you understand physics well enough, you ought to be able to predict the state of that object at any time in the future. I’m thinking in a particularly classical way, but that is generally true. And so the observations of the cosmic microwave background told us the state of the universe shortly after the Big Bang, and then with our brand new shiny cosmological model, we could use that model to then propagate forward or extrapolate to the present time how fast the universe ought to be expanding. It’s really, I would say, a prediction. It’s not a measurement. The measurement occurs shortly after the Big Bang, and then the model is used to make a prediction of how fast it should be expanding today.

0:26:10 SC: And the specific thing that you’re measuring is not just the amount of radiation in the microwave background, but the temperature of it and how that is different from point to point in the sky when you look at it, right?

0:26:21 AR: That’s right. That’s correct. So you’re measuring the spectrum of fluctuations, and so you see this kind of mottled-looking map and physicists then take what we call Fourier transforms to measure how much of that spottiness on different scales exists. And because this is kind of a ringing oscillating fluid, shortly after the Big Bang, by comparing the actual measurements to models that predict what those measurements should look like, we essentially set the what we call the free parameters of the cosmological model, and to then use that calibration of the model to propagate the story forward to the present.

0:27:02 SC: So this is, it seems to be, if I pretend to be an outsider for a second, this also seems to be a bit of a ladder of some sort. In other words, when we try to measure parameters using the cosmic microwave background, it’s not like the CMB is holding up a sign saying, “Here is the density of matter,” or anything like that. There’s these temperature fluctuations, and we have a very, very theoretically motivated set of predictions for what those fluctuations should be, depending on the temperature, depending on the parameters, and then we fit. And it’s sort of a back and forth.

0:27:38 AR: That’s correct. And the way the game is normally played is you start out and say, “What is the fewest number of parameters or the simplest physics I can use to explain the data that we have?” And so that’s what allows you to fit this model at early times. One thing very important for people to understand is the universe looked very different shortly after the Big Bang. It was mostly matter and neutrinos and radiation. The universe then looked very different at late times. It’s mostly dark energy and a fair bit of matter, neutrinos become unimportant. And so you fit with the simplest physics you can this model and then cross fingers, hope that that is the complete story and then use that model to make a prediction about the present. Of course, one of the challenges in this is 95% of the model as we know it today is dark. It’s dark matter and dark energy, and we don’t really know much about the physics of dark matter and dark energy. And so what we do is we take the most vanilla versions of those, in the case of dark matter, we say, okay, it looks like a particle for various reasons. We don’t know much about the interactions of that particle. We don’t know how stable that particle is. We don’t know if it decays, but let’s assume none of those things.

0:28:57 AR: And in the case of dark energy, while we don’t understand it, we say, let’s say that’s static and fixed just like the cosmological constant, and then we use that most vanilla version to make this prediction. And just to jump right to the punchline, what we see, and it’s pretty remarkable right now, is that we do not see agreement to the precision of the measurements.

0:29:19 SC: Right, that’s the most recent wrinkle that we want to dive into, but it’s a tough line to walk down because I want to care about and really think carefully about the most recent tension. But also there’s a lot of agreement, and it’s something where if you’re not in the game, I think a lot of people say, “Well, look, you’ve invoked this dark matter stuff, this dark energy stuff. I’m not surprised it doesn’t fit at all. Maybe you just don’t know what you’re talking about.”

0:29:49 AR: Right. Yeah, no, it’s certainly a give-and-take sort of process. There are ways or there are other measurements, I think, that confirm and corroborate each of these different pieces. In the case of dark matter, we see the rotation curves of galaxies are flat, which is the way we say that galaxies are really spinning way too fast if there weren’t lots of dark matter in them. And we see galaxies whizzing around clusters of galaxies, much too fast for those galaxies to stay connected to those clusters were it not for dark matter in those clusters. And then in terms of dark energy, we see… We have other ways of gauging distances in the nearby universe that have nothing to do with supernovae that also show the universe to be accelerating. And so this is a multi-piece puzzle, and the cosmological model we have today has sort of survived all of these various measurements. It’s the best thing that we have now, but it doesn’t necessarily mean it’s complete.

0:30:53 SC: Yeah. I guess the thing I want to emphasize is, and if you agree with me here, we would love to be able to say the model was wrong, right? That’s the thing that makes us happy is I think that a lot of people have the idea that we somehow have a vested interest in the success of dark matter and dark energy being simple and whatever, like we have bought stock in it. But as scientists, we’re most excited when it’s completely wrong.

0:31:21 AR: Well, that’s right, because I would say it depends a little bit about your model. If we thought we understood absolutely everything about the model, we might be a little horrified if the model didn’t work somewhere. But when we have a model that has such large swaths of ignorance, really, then seeing the model not work somewhere offers the best opportunity or really clue we could potentially have to the physics that we’re missing in the model. And so I agree with you. In the current context with so much dark matter and dark energy, if something didn’t fit, that would be fantastic news.

0:31:55 SC: Good. So let’s get into the modern tension, but first I just want to sort of summarize what we mean when we say “the model.” So we have the hot Big Bang. We had these parameters. Why don’t you put it into your own words?

0:32:07 AR: Sure, we have the hot Big Bang, and then we have different constituents in the universe. We have particles called neutrinos that have very little mass. A lot of them, they whiz around. They act a lot like light or energy. But as I said, they have a little mass, so they don’t quite travel at the speed of light. And physicists have certain descriptions of how to explain those. We have, of course, radiation. We have this dark matter. We have dark energy. And as the universe expands, these different components will dilute in different ways. What else do we have? We have the basic laws of physics as we understand them. We have a theory of gravity, general relativity that is sort of embedded in all this because we always have to question whether we have the right story there as well. So we say the cosmological model, but we really mean certain constituents in certain amounts and certain physics.

0:33:08 SC: And we have confidence that the dark matter is not just ordinary matter we haven’t seen yet.

0:33:13 AR: That’s correct.

0:33:14 SC: Where does that come from? [chuckle]

0:33:15 AR: Yeah, so I guess that has to do with… Depending on what the density of matter is in the early universe, it tells us, through a process called nuclear synthesis, when the universe was very hot and dense, how much fusion will go on and how much of certain primordial elements you will produce. For example, helium is a great example. And so if all of the matter in the universe was normal or what we call baryonic matter, you wouldn’t produce the right amount of helium. And so that’s a critical piece of this story, is to have some matter that is not ordinary. In fact, most of it is in what we call non-baryonic or some form that you wouldn’t recognize from your high school chemistry class in the periodic table of elements or the list of basic particles. And so that’s very important.

0:34:10 AR: Back 20-30 years ago, people used to argue whether those were particles or whether those were giant bricks or macroscopic objects. And well, I would say there’s still some debate about that. For the most part, we’ve ruled out that they’re macroscopic objects because we don’t see enough other effects from those macroscopic objects, something called microlensing, when one of those bricks or macroscopic objects would pass between us and a star, it would produce a kind of change in the brightness of that star. And we don’t see enough of that to find that much of it is in that form.

0:34:47 SC: I remember when the LIGO results first came out and we discovered that there was gravitational waves coming from binary stars. Were you one of the people who wondered out loud whether or not big black holes could be the dark matter?

0:35:00 AR: Yes. Yeah, I was. Really, Marc Kamionkowski really spearheaded that. I think I contributed a lunch-room conversation to that paper. However, I think it’s really intriguing. Today, LIGO is finding these very large black holes that can be like 60 times the mass of the sun, and what we know, it’s very, very rare for stars that large to form or particularly for the remnants of stars that large to exist. So there’s some kind of process that may be producing bigger, heavier things from smaller things glomming on together. And so one could imagine that that’s been a recent process of massive stars coming from black holes that are maybe one, two, four solar masses glomming on to each other in the centers of clusters of stars, or it could be the kind of snowball effect from little things starting off shortly after the Big Bang that were black holes and then glomming on over a much longer period of time. And so that’s an active area of research. It’s something I think that’s very interesting.

0:36:09 SC: Do you still think it’s an open possibility that the dark matter is black holes of some form that are left over from the early universe? Or do you think that…

0:36:16 AR: Yeah, I haven’t followed this story as closely as I would like, but I’ve seen a fair bit of back and forth. There are other constraints that make it difficult for it to be a lot of the dark matter; it could be, I think, a small amount of the dark matter. And I think the jury’s still out at that level.

0:36:29 SC: Okay, okay, good. Alright, so the final sort of preparatory question is, when we talk about the standard cosmological model as you described it, there’s expansion, there’s dark matter, there’s ordinary matter, there’s dark energy. What about inflation? This is an idea in the very early universe, super rapid expansion. It’s certainly a very popular idea. In your mind, do you count that as part of the standard picture or as a more speculative add-on?

0:36:55 AR: Right. I guess, and now, I’ll have to remind myself a little bit after this conversation. I don’t usually think or deal with the universe much around the time of inflation. However, it does set initial conditions for the standard model and some of the parameters in the standard model that effect the cosmic microwave background do come from inflation. I think… My understanding usually is inflation is a generic enough phenomenon that they don’t necessarily prove that inflation occurred at this point. But it is part of the setting of the stage for what comes next in the observations of the cosmic microwave background.

0:37:38 SC: Okay, but then… So I take it that you would say that if we find out that inflation was not right, we would still say the general Hot Big Bang model could be completely correct. In fact, I think the general Hot Big Bang model is more or less just correct, right. That’s not still up in the air to debate.

0:37:52 AR: That’s correct.

0:37:53 SC: Okay, so nevertheless… Good, so with all that groundwork laid, there seems to be an issue. You’ve kept on measuring the Hubble Constant, the expansion rate of the universe. You sort of started out measuring it and then you showed that it was…

0:38:09 AR: It was a terrible mistake, actually, to keep making measurements. One should certainly stop when everything looks good.

0:38:15 SC: There are laurels, you should rest on them. Why don’t you take this advice? Come on.

0:38:19 AR: That really was a blunder.

0:38:22 SC: And we also have this wonderful picture where there’s multiple sources of data and inference from whether it’s supernovae or galaxies or the microwave background or whatever, and what we would hope is that we sort of, with improved precision, narrow in on the right answer. And it seems that with improved precision, we’re finding that it doesn’t quite fit as nicely as we would like.

0:38:44 AR: That’s right. So I would say in broader brush strokes, the early 2000s was a period of great harmony in cosmology where various techniques, at better precision than before, but not as good as precision as we had now, were pointing to the same gross order picture, this Hot Big Bang model with a large dollop of dark energy and a smaller dollop of dark matter, and this was all glorious and we were all patting ourselves on the back. And then, as I said, we made the mistake of continuing to improve the precision of all the measurements, and then we started to see these tensions, really, I would say, between the early universe’s measurement and predictions coming from the model to what the state of the universe should be today versus a more direct measurement of that present state of the universe. And these have come up in a couple of ways, one a little better known than the other. One is measuring how fast the universe is expanding, the Hubble Constant, and the other is measuring how clumpy the universe is today. And both of these seem to be in some level of tension or difference.

0:39:57 SC: So good, let’s dig into both of them one at a time. So the Hubble Tension, as it’s been called, roughly speaking, two different ways of measuring the expansion rate, they disagree.

0:40:08 AR: Right. But as I said, the key thing is, one is a measurement and then the use of a physics model to make a prediction, and the other is a more direct version of that measurement. And so when you start at the early universe and you make those measurements and you use the model, you predict the universe should be expanding at about 67 plus or minus 0.5 kilometers per second per megaparsec. And when you make those measurements locally, while there are various methods, and I would say they get a full range of about 70-75, all of those methods are coming out higher than the early universe. And I would say the average is sort of around 73 plus or minus 1.5. So we’re looking at somewhere between 4-6 error bars or standard deviations of separation now between the average of those late universe measurements, the present measurement, and those early universe predictions, so much so that I really can’t point to any one thing that could be an error that could explain this difference, because there are so many independent methods that are used at both ends.

0:41:29 SC: So let’s just narrow it down even a little bit more. What is your favorite way of directly measuring the expansion rate today?

0:41:35 AR: Well, my favorite way is, must be the one I work on because you have choices in this world, and so those are the ones I’ve chosen to work on. And so the Type Ia supernovae, I think, are without a doubt the best way we have of measuring long range differences and these pulsating stars called Cepheid variables. They’ve been around the longest, going back more than 100 years when Henrietta Leavitt measured them in our nearby neighbor, the Large Magellanic Cloud. They are also about the most luminous star you could see in a distant galaxy, and they have a wonderful property, which is that they blink, they pulsate. And the period with which they pulsate, which can be from days to months, has a very tight correspondence to their luminosity. It’s because these are supergiant stars and as you might imagine, something far larger and more massive, picture this very large star whose orbit probably goes out to Neptune or something. And imagine how long it would take to have a star like that compress and expand like an accordion. And the bigger it is, the longer it’s going to take by the process of gravity.

0:42:50 AR: And so the very big ones are the more luminous ones, and their period tells us that, the fact that it takes a very long time for them to undergo one of those pulsations, whereas the smaller and less luminous ones might only take a day or so. And so, we look at a nearby galaxy that hosted a Type Ia supernova with the Hubble Space Telescope, and we can pick out individual luminous stars and we monitor that galaxy maybe a dozen times over the course of a couple of months and we identify these Cepheid stars by their changing brightness. They might double their brightness over the course of a couple of months, and we’re able to pick them out and to compare them to their nearby brethren, many of which we can now measure parallax to directly. And so it forms a pretty simple measurement chain, parallax to the Cepheid variables in the Milky Way galaxy, observe those with the Hubble Space Telescope, observe their brethren in nearby galaxies that hosted Type Ia supernovae, and then observe those Type Ia supernovae further into the expanding universe.

0:44:00 SC: Cool, so that’s the three-step ladder that you said that we’re now able to do. So Henrietta Leavitt was stuck just sort of picking out Cepheids in the Large Magellanic Cloud and presuming they were more or less the same distance away, but now we can basically use parallax to measure the distance to some Cepheids directly. Is that right?

0:44:16 AR: That’s right. Henrietta Leavitt made this very important realization that because all the Cepheids she was looking at were in one location far away from us, but in one location, that she noticed the ones that were more luminous had the longer periods. And so this allowed her to recognize that they didn’t just look brighter, they actually were brighter. They were more luminous ’cause they were co-located with other Cepheids that were fainter and had shorter periods.

0:44:44 SC: And then we go from Cepheids in other galaxies to supernova because there are some supernovae that are in the same galaxies as the Cepheids. How many Type Ia supernovae have we observed with Cepheids in their galaxies?

0:44:58 AR: So a Type Ia supernova is a pretty rare phenomenon. There’s only one in a galaxy like ours every 100 years, and so most galaxies we know of, we have not seen a Type Ia supernova in it. There’s only really some handfuls that we have nearby. However, most galaxies do have these Cepheid variables, so we do the process sort of in reverse. First you find the galaxy with a supernova, then you go back and look for Cepheid variables in these galaxies. With the Hubble Space Telescope and particularly with its newer instrumentation, which has only been available in the last 10 years or so, we can do this out to a distance of about 40 or 55 megaparsecs. And in that volume there is a good Type Ia supernova about once a year, maybe once every two years.

0:45:47 AR: And so we are just this year completing, I think, a complete sample of what I would call all of the Type Ia supernovae in the last 30 or 40 years that you can calibrate with Cepheid variables, and so that’ll be about 38 objects total. And the precision of each of those Type Ia supernovae is about 6% in distance. And so if you have 38 objects that you can measure individually to 6% in distance, that gets you close to about 1% overall, and that is sort of the holy grail of measurements of the Hubble Constant is to get to a 1% precise measurement. Our most recent effort of a few years ago was more like 2%. And this may sound like small differences, but when you’re trying to understand the cause of the tension, improving the measurements is really one of the better ways we have.

0:46:43 SC: I mean, again, when I was a graduate student, we didn’t know whether the Hubble Constant was 50 or 100, [chuckle] and so now we’re arguing whether it’s 72 or 73. So that’s far…

0:46:51 AR: Right. This is much more interesting, because I would say when you and I were graduate students and people were measuring whether it was 50 or 100, they were measuring whether it was 50 or 100 in the same nearby universe, often with some of the same data. And so you knew both things couldn’t be true. What’s so interesting right now is we’re talking about a difference between what we see in the nearby universe or what we infer from early times, and it is possible that both those measurements are true and that we’re learning something isn’t quite right about the model. We don’t know that for sure. That’s why we continue to improve the measurements, but that’s why it’s a fundamentally different and potentially much more exciting possibility than the “less filling tastes great” argument.

[chuckle]

0:47:37 SC: Well, it was hilarious with the 50 versus 100 is that nobody thought it was 70 or 75, even though that was the average and that’s what it more or less turned out to be. Let me, before we keep moving on, you mentioned the Hubble Space Telescope. And number one, how amazing is it that that guy is still going, running strong? And number two, is the next generation, is the James Webb Space Telescope going to make your job much easier?

0:48:03 AR: Yeah. Well, first of all, it is amazing that Hubble’s still going, and it’s really a testament to NASA and the space program, the astronauts, because of being able to go back up and service it. If the astronauts had not been able to do that a number of times, we would not have the telescope now, for sure, but they also brought it very much up-to-date. And so one of the wonderful things when you have a complex piece of machinery like the Hubble Space Telescope is things break. Particularly early on, things have sort of infant mortality, and so if you can service it for a while, if you could spend a while replacing all the parts that broke easily, then what you get by the end is something that’s quite robust, quite sturdy.

0:48:45 AR: And so we have not only a pretty up-to-date telescope because it was serviced 10 years ago with modern instruments, but we also have one that’s quite robust because we’ve replaced all the bad parts. And so that’s what makes it quite incredible today. The next telescope, the James Webb Space Telescope, will allow us to see further into the past. It’s not particularly optimized, I would say, for the projects that I do, Hubble really is better for that, but I think that there will be other kinds of measurements you can make with the James Webb Space Telescope that shed light on this issue, sometimes in expected ways, sometimes probably in unexpected ways.

0:49:23 SC: Is Hubble going to be retired? Is it going to keep going even when we get the James W Webb up there?

0:49:28 AR: Yeah, well, you know, there’s no retirement ceremony where we hoist its jersey into the rafters and put it in the Hall of Fame. We’re going to keep using it [chuckle] until so many parts are broken, and that’s, as I said, not the case now, that it isn’t particularly useful or costs too much to continue to operate. At some point, its orbit is sinking, and so it begins to touch the atmosphere which causes its orbit to sink even more. At some point, it will re-enter or NASA will have to attach something to it and drive it into the ocean, because it’s too large to allow it to just fall.

0:50:07 SC: And that’s the price you pay for having it be serviceable. If it’s in a near-Earth orbit, then it will decay.

0:50:11 AR: That’s right, absolutely.

0:50:13 SC: And the JWST is not going to be serviceable. It’s going to be way the hell out there.

0:50:17 AR: Yeah, it’s going to be way the hell out there, much further away than the moon is at a point, the second Lagrange point, and that of course makes it even more critical that it works very well from the beginning because astronauts are not going to be able to go out there.

0:50:31 SC: Okay, very good. Thank you for the technological update here, but so we had this nice relatively clean three-step Hubble distance ladder. Let me just ask, what are the prospects for a lurking systematic error, like Cepheid variables in the Milky Way are all weird, and that’s the very first step on the ladder?

0:50:51 AR: Yes, so that’s a great question. So I gave a kind of a simplified version of the ladder, but we have now, I would say, three pretty equally good ways to calibrate the luminosity of Cepheids. One is actually the parallax measurements in the Milky Way. We also see those Cepheid variables in the Large Magellanic Cloud, and there are other geometric techniques that allow us to measure the distance to the Large Magellanic Cloud, other kinds of systems called detached eclipsing binary systems that calibrate the Cepheids to about 1.2%. There are also lasers in Keplerian motion around a supermassive black hole in a galaxy called NGC-4258, and we calibrate the luminosity of the Cepheids in there. So we have about three completely unrelated methods for making those measurements, and they all give a very consistent result that is consistently in tension with the early universe. And so that’s why I say it’s, we’re having a harder and harder time getting out of this problem by blaming systematic errors. We continue to look carefully for them diligently. It’s funny, we talked about that discovery 20 years ago about the acceleration of the universe.

0:52:09 AR: Ironically, I would actually say this is an easier measurement than that was because 20 years ago, we had to do our best to figure out and conclude that distant Type Ia supernovae when the universe was very young are the same as we see today. In this case, the measurements I’m talking about are all done essentially today. They’re all done nearby at low redshifts, and so many things are easier about this measurement, and so it’s been harder to understand some kind of systematic error that could be causing this.

0:52:40 SC: So we have a bunch of different, independent, fairly robust-seeming ways of measuring the local, nearby expansion rate, and we get numbers like 72 or 73, and then we have the cosmic microwave background that implies something like 67. This is a difficult question, I know. Is it possible to give an intuitive understanding of how you determine the expansion rate of the universe today. ’cause you’re not trying to discover the expansion rate of the universe when the microwave background was formed. You’re using data from the microwave background to infer the expansion rate today. So how is that supposed to work?

0:53:17 AR: How is it that you infer an expansion rate today from the early universe?

0:53:21 SC: Yeah.

0:53:24 AR: I think the easiest picture for anybody to have is imagine that we are measuring essentially the expansion rate of the universe shortly after the Big Bang at that time, and then we have a kinematic model, that’s essentially what the model in the form of a well-known equation called the Friedmann equation allows us to propagate that forward in time to the present time. And so it’s a prediction; however, there are checks we get along the way. For example, in that calculation, we get to check along the way, are distant supernovae or other tools called baryon acoustic oscillations, are they showing us an expansion history that is consistent with that fit to the Friedmann equation along the way? And the answer is yes. And so that gives us a lot of confidence in this process. However, there’s, I would say at this point, in my mind one of the bigger flies in the ointment is, do we have the right description of the universe, not just today, but before the cosmic microwave background light is emitted?

0:54:28 AR: That is, do we have the right model of the universe shortly after the Big Bang when it would have been dominated by neutrinos, radiation, dark matter, and it would have been so hot and dense that it makes you think about interaction of dark matter or interactions of dark matter with dark energy or with neutrinos in a way that we wouldn’t recognize today because now the universe is very diluted because of the expansion. And so it’s a very different universe shortly after the Big Bang, are there any kind of other inhomogeneities? Are we missing a particle? I’m an astronomer. I don’t usually think about particles, but when the universe is very young, it’s all particle physics. It’s all a giant accelerator. And if you have another neutrino, for instance, it changes the proportions of energy in the early universe and changes our understanding of how fast it would expand at early times, and it changes essentially the calibration of the cosmic microwave background.

0:55:28 AR: Just like we calibrate the luminosity of a supernova, that’s our standard candle, they have a standard ruler, really, which is the physical size of a fluctuation in the plasma of the early universe, which can travel from or expand from the time of the Big Bang to the moment when the universe becomes transparent, when the radiation can escape and travel to us. And if we do not understand the expansion at those early times, for example, if there’s another neutrino in the universe, it causes more energy to be in what we call relativistic-like particles, it would increase the expansion at early times and thin out the universe faster, causing it to become transparent earlier, which changes the size of this sound horizon with the distance that this sound wave can travel up until that time. And so that’s something people have been taking a look at recently is are there ways to change the cosmological model, a piece, a wrinkle, something we’re missing either in the physics or the components of the model, but at early times that could bring these two things into accord?

0:56:43 AR: And for example, one possibility people have looked at is what’s called early dark energy, another episode of dark energy, so if you’re counting at home, this would be the third episode of dark energy. There is the one that we think gave rise to inflation, there’s the one that accelerates the universe today. This would be a field-driven dark energy that would exist shortly before the Big Bang and would push the universe apart faster. So those are the kinds of things people look at.

0:57:14 SC: So okay, good. So let me just sort of see if I can absorb all this. So we have the direct measurement of the expansion rate today. That’s the 72 or 73. We have this fitting to the big old model and using the cosmic microwave background that seems to imply that we should measure the Hubble Constant today to be 67, so those disagree. And my impression, both from having read things… I’ll confess. Despite having a history as a theoretical cosmologist, I’ve not followed the latest wrinkles in this game, so I’m relying on you here. But my impression both from reading things and from listening to you right now is that the ways that we could tweak the models to bring these discrepancies into alignment once again are focusing more on the early universe than on the current universe. You might think, “Well, the CMB predicts something. We disagree with it today. Therefore, let’s change today.” But people seem to be saying that that’s harder to do than to add some new things to the early universe.

0:58:16 AR: Right, it’s not so much that it’s harder to do theoretically. It’s that it’s harder to do from the data is really reining us in at this point. If you would have told me about this problem 20 years ago, shortly after discovering dark energy, our first statement would have been, “Oh, so it’s not a cosmological constant. It has an equation of state, the ratio of its pressure to its energy density, which is more negative than minus one,” that 20 years ago would have bothered you very much because…

0:58:46 SC: Very much. [chuckle]

0:58:47 AR: I think you… [chuckle]

0:58:48 SC: Still would. Still would bother me.

0:58:49 AR: That would be very unlikely, but observationally, we would have said that. Now with additional observations of high rates of supernovae, baryon acoustic oscillations and other techniques, we’ve seen, oh, no, the equation thing is actually very close to minus one within 5% or so. And so therefore, the late-time behavior of the universe is much like the standard model. And so that’s why, as I said, we turn our attention to another place. The model really comes in in two critical places in this story. One is before the radiation from the Big Bang comes out to us, and the other is in the late universe. And so if the late universe is more off the hook, then we look at the early universe.

0:59:29 AR: Now, it’s not easy in the early universe either because the spectrum of the cosmic microwave background has been measured so exquisitely that if you try to monkey around, you throw in a neutrino willy-nilly, it will produce other changes in the spectrum that are not allowed. And so we’re reined in, I would say, less than in the late universe in terms of creativity space, but still quite reined in. And so it’s hard. It is, I would say, hard, both from a theoretical standpoint and also from a systematic error standpoint. I might say at this point that there’s not right now one sort of silver bullet that comes in and explains the whole thing, but there are interesting theoretical ideas that get close, but are what you might call fine-tuned a little bit. There are suggestions people have made of systematic errors that might knock out one observation but not other observations, so then you need multiple, maybe a conspiracy of systematic errors. And so that’s why it’s the Hubble tension.

1:00:36 SC: Right. [chuckle] Well, maybe before getting deeper into the ideas, maybe just a personal, a Bayesian prior or a credence that you can offer us. What do you think the chances are that this discrepancy will at the end just be a systematic error or some mistake in the observations versus being new physics or something exciting?

1:00:57 AR: Yeah, you know, it’s really hard. I’ve learned to not try to guess. [chuckle] My favorite example from teaching Astro 101 is the parable of Neptune and Vulcan, or maybe I should say Uranus and Mercury. Uranus was not traveling as it ought to have been, and astronomers speculated that there was another planet that was tugging on it, which turned out to be Neptune, and in fact, those calculations allowed them to look exactly where it was. And so in that case, it was actually a new object to see. In the case of Mercury, Mercury was precessing in its orbit, which was recognized a couple of decades later, and people tried the same trick. “Ah, I’ve seen this problem before. There must be another planet between Mercury and the Sun,” which they named Vulcan ’cause it would have been so darn hot there. And they looked and looked for it for decades and never found it, and it turned out they had the wrong physics, the wrong theory of gravity.

1:02:02 AR: So you might call Neptune the systematic error, and you might call general relativity that came from this the new physics. And sometimes you could look at the exact same discrepancy and it’s one or the other, and your intuition fails you. So I don’t know. To me, this is one of the most interesting things going on in the universe right now. I think it is something interesting, and I don’t consider a simple systematic error as interesting. So I think it is something that will tell us something about the universe around us, but whether that’s local about the astrophysics of objects in the universe or whether it’s something more about the actual physics like gravity, I don’t know.

1:02:48 SC: Okay, so you’re weaseling out. I get it, no problem. You have the right to [chuckle] do that. Well, let me ask this. You mentioned that there were these proposals out there, hardworking young theoretical cosmologists making up crazy ideas for the early universe.

1:03:00 AR: [1:03:00] ____.

[laughter]

1:03:03 SC: My particular style for doing cosmology is less, let’s propose a particular new particle or whatever, and more, well, what’s the general kind of thing we need. Is there a thing that you can say like, if we just made the universe expand a little faster because there was more energy density at early times, that solves all of our problems, or is it more subtle than that?

1:03:31 AR: Yeah, you know, at this point we’re tightening, I would say, the noose at both ends of space and time, and we’re looking for… Well, it looked like a just-so story. Well, sometimes that happens. You know, Hoyle seeing the resonance of carbon was a just-so story because it had [1:03:51] ____ at some point. It could be dramatically different physics and we just don’t have the right picture right. It could be nature’s so cruel to us, it’s thrown us three different systematic errors [chuckle] that all go in the same direction.

1:04:06 SC: I mean, you laughed, but it could be. That’s a real possibility.

1:04:10 AR: You know, I mean, you’d be crazy not to say that. These are measurements. These are observations. Life is hard. That those systematic errors could be either in the late-time or early-time universe or of both of them. I know smart people who will say, “Oh, it’s just like in the past, was it 50 or 100? It was in between.” And so they’ll say, “Oh, is it 67 or 73? Well, it’s 70. It’s in between.” That’s not very satisfying because, for the most part, saying that there’s a tension between 67 and 73, you don’t escape that by saying, “Well, the answer is in the middle because both ends don’t like that very much.”

1:04:50 SC: Now you have two tensions.

1:04:52 AR: Sorry?

1:04:52 SC: Now you have two tensions instead of just one.

1:04:54 AR: That’s right. You have two tensions, so it might seem clever to sort of conclude that, [chuckle] but we’re not really interested, believe it or not, in the value of the Hubble Constant. We’re interested in why we don’t get a match between these two ways of looking at things. That is what potentially tells us more about the universe. Just if aliens arrive and just told us the answer and whatever they said it was, that wouldn’t necessarily explain what is going on. So this is not a quest to measure a number, this is a quest to do an end-to-end test of the universe and understand why it’s failing or we’re failing.

1:05:28 SC: You did tease us a little bit at the very beginning with mentioning that it’s not just the Hubble measurement that’s in tension, but there is this problem with the lumpiness of the universe as well.

1:05:38 AR: That’s right. So there’s another parameter, which often goes by the catchy name of Sigma 8 and…

1:05:43 SC: Sexy.

1:05:44 AR: And it measures the clumpiness of matter. And so likewise in the early universe, because we know the state of the universe shortly after the Big Bang, and then we know how it dilutes due to expansion, we should be able to predict how clumpy it should be today. And then we have techniques which are also hard but are powerful in measuring how clumpy is, either by measuring this phenomenon called loop lensing, where we measure how clumps of matter distort the shapes, the apparent shapes of galaxies, or by measuring the motions of galaxies, extra motions due to not the expansion of space but matter, the clumpiness of matter, and those two at this point disagree with each other at… I guess the most recent number that I’ve seen is sort of 3 sigma from each of those, the lensing method and the peculiar velocity method. And in the same direction, that the late universe measurements are in a different direction in terms of size to the sigma 8 number as multiple early universe measurement.

1:06:43 AR: So clever people are trying to think of… Can they think of one physics story that would explain both of these tensions. I think that’s turned out to be equally difficult, but the reductionist in all of us sort of wants to think if there is a physics answer, it would explain both.

1:06:58 SC: It would be simpler, yeah. I mean, this is an entirely unfair question, but this is the benefit of being the questioner rather the questionee, is it possible that there’s a big thing that we’re missing, not like just if it’s another particle or something, is there something that we’re just misconstruing about the nature of the universe that would help explain this kind of thing?

1:07:16 AR: You know, I go back to gravity in a way, because if somebody was looking at the precession of Mercury and somebody whispered in your ear, “It’s new physics,” you would go, “Oh, it must be a small perturbation in Newton’s gravity, right?” We pretty much get everything right, this just was one little detail left. And that is completely wrong. As you know, general relativity is nothing like Newtonian physics, it’s not a wrinkle. And so the moral of the story is it would be very difficult to predict. We can explain an enormous number of phenomena in the universe with the standard model, and there’s only, as I said, at present a couple of these tensions, but are these the threads in the sweater that you pull on them and the whole sweater unravels, or are they little threads that you pull them and you go, oh, those were just some loose threads on the sweater. And it’s very difficult to guess. You might have a better, more intuition about that than I do.

1:08:10 SC: Well, I’ll tell you my intuition, which is that as someone who’s tried to do this, changing gravity in such a way that it fits the solar system and the binary pulsar and gravitational waves and all that, but is different for cosmology, is just not only hard because we’ve tried it and it’s been difficult, but also not at all what you’d expect from the philosophy of effective field theory in something like that. Like in the weak regime where the whole universe is there, we should be on firm ground, but it’s in the nature of large conceptual mistakes that you don’t know you’re making them while you’re going on, so I’m in favor of keeping an open mind.

1:08:46 AR: That’s true, and not only that, when you start talking about things like dark energy and the cosmological constant or even inflation, and you talk about what is natural in our current understanding of gravity, you run into real trouble in terms of understanding the size and scale of, for example, the cosmological constant. So it’s one thing to say, well, I just have these few tensions, these loose threads in a sweater, and it’s another thing to also recognize, and I expect some enormous newer understanding to come along because I also have this other enormous lurking problem, which is the ignorance about dark energy. Why is lambda, as we call it, so small? And so all of these things are kind of in the backdrop, so I don’t know whether… You know, is the problem these loose threads on this sweater or the fact that we don’t know how to make a sweater at all?

1:09:38 SC: Right, and I think the thing that is important is not only that the vacuum energy that we need to invoke to fit everything is small, but also the coincidence problem, that it’s approximately the same order of magnitude, it is the same order of magnitude as the matter density, which leads people to wonder about the anthropic principle and the multiverse, but also might lead us to say, well, we just don’t have a good idea what’s going on right now. And I think even as I say that and the words escape my mouth, I feel the need to say, but do you know what, that idea fits a huge amount of data, so don’t just dismiss it too lightly, it’s a typo.

1:10:14 AR: No, that’s right, that’s right. In any field, and I think this is a good field, the measure of how rich it is is you ask for its top 10 list of problems and you hope that they’re not all tiny it’s an interesting field. And this is that kind of field where there are these enormous problems, but then there are these sort of little clues, and so the question is, do they connect. I mean, it’s not anything goes, as you say, there are a great number of things that we know quite well and have measured quite well and work quite well with the theory, but then, like the analogy of Newtonian gravity to general relativity, is this just a special case of a deeper theory, that’s what we don’t know. And so clever people, smarter than me, for sure, look at all of these clues and try to put it all together.

1:11:05 AR: One interesting question people often ask me is, how close are we to a theoretical breakthrough and to… When you think about Einstein and general relativity, what were the clues over the year or two before, you know, did you know we were almost there? Is there markers of progress along the way before giant theoretical breakthroughs come through? I don’t think so.

1:11:25 SC: Yeah, that’s a tough one. Of course, you can’t predict the next theoretical breakthrough, so that is a one to wisely, safely pass on when people ask you that question. So instead, let me wind things up by asking you a question you’re much more equipped to answer: What are the things we should be looking at next, like be the observational cosmologist here, what are the tactics we can use to actually learn more about the phenomenon here?

1:11:50 AR: Right. So there are going to be a great number of new and highly precise measures of the late-time expansion history of the universe coming from all kinds of new facilities that will tell us if indeed there is a component of this story coming from from some misunderstanding there or whether we can rule that out completely. At the same time, there will be new observations of the cosmic microwave background in greater detail and other scales of fluctuation and polarization that can look for extra features in the cosmic microwave background that would be predicted if there is some other component of the universe at early times. And so that’s important. There will be more measurements of the expansion rate of the universe, there’s this facility, LIGO, that measures gravitational waves, it’s going to be measuring objects that we call standard sirens, they’re like standard candles only in gravitational waves instead of light that can be used to independently measure the Hubble Constant at late times, look for any sources of systematic error.

1:12:56 AR: And so all of these will come together and to… I think over the next decade, I’d be surprised… I mean, most of the time when you’re looking at a problem if it’s been around for 50 or 100 years, it’s going to be around a lot longer. But the Hubble tension’s maybe only been around five years, so maybe the next 10 years will be… We’ll really solve it, I hope so.

1:13:17 SC: Yeah, I actually was, I was one of the ones who coined the term standard sirens, that was one of my contributions to cosmology. I don’t make major breakthroughs myself, but I’m friends with a lot of people who do, so I helped change the way that we talk about them. But also maybe one last food for thought thing is the difference in sort of style of what’s going on right now versus what went on 20 years ago with the acceleration of the universe. I mean, with that, it was a big, important, huge surprising discovery. But it was pretty quick. Like you made it and people said, oh, yeah, okay, this is what’s happening. And in part because we had a theoretical explanation ready to hand, and this one is much more creeping up on us, the discrepancy has been there for a while and it’s growing, and if it’s real we’re still not quite sure what to do with it.

1:14:09 AR: Yeah, it is an interesting contrast, and I could almost argue why there’s [1:14:13] ____ of that. You know, 20 years ago, we said, oh, you know, the universe is accelerating, could that really be? And right away people said, yes, because there’s a cosmological constant. Aha. We have a ready explanation. And yet, as you know, that explanation comes with a huge number of problems as well, is that really an explanation or is that kind of a place holder? There’s math and some rudimentary physics, but not a really detailed understanding, and so it’s hard to say whether, maybe we don’t have that sort of idea now, but as I said, I’m not sure that that idea that we had 20 years ago is the final story there either.

1:14:53 SC: Yeah, I’d like to think that if this podcast has any function it’s to be listened to by young people who are still in their formative years and they’ll be inspired to solve some of these difficult problems we keep talking about, so maybe, maybe someone in the audience, somewhere a Mindscape listener will go solve the Hubble tension problem. That would be a good outcome.

1:15:11 AR: Yes.

1:15:12 SC: Alright, Adam Riess, thanks so much for being on the Mindscape Podcast.

1:15:13 AR: Alright, thank you, Sean.

[music][/accordion-item][/accordion]

Scroll to Top