Book Review: We Have No Idea

I have no idea how I’m going to review this book.

Ok fine, I have some idea.

Jorge Cham writes Piled Higher and Deeper, a webcomic with possibly the most accurate depiction of grad school available. Daniel Whiteson is a professor at the University of California, Irvine, and a member of the ATLAS collaboration (one of the two big groups that make measurements at the Large Hadron Collider). Together, they’ve written a popular science book covering everything we don’t know about fundamental physics.

Writing a book about what we don’t know is an unusual choice, and there was a real risk it would end up as just a superficial gimmick. The pie chart on the cover presents the most famous “things physicists don’t know”, dark matter and dark energy. If they had just stuck to those this would have been a pretty ordinary popular physics book.

Refreshingly, they don’t do that. After blazing through dark matter and dark energy in the first three chapters, the rest of the book focuses on a variety of other scientific mysteries.

The book contains a mix of problems that get serious research attention (matter-antimatter asymmetry, high-energy cosmic rays) and more blue-sky “what if” questions (does matter have to be made out of particles?). As a theorist, I’m not sure that all of these questions are actually mysterious (we do have some explanation of the weird “1/3” charges of quarks, and I’d like to think we understand why mass includes binding energy), but even in these cases what we really know is that they follow from “sensible assumptions”, and one could just as easily ask “what if” about those assumptions instead. Overall, these “what if” questions make the book unique, and it would be a much weaker book without them.

“We Have No Idea” is strongest when the authors actually have some idea, i.e. when Whiteson is discussing experimental particle physics. It gets weaker on other topics, where the authors seem to rely more on others’ popular treatments (their discussion of “pixels of space-time” motivated me to write this post). Still, they at least seem to have asked the right people, and their accounts are on the more accurate end of typical pop science. (Closer to Quanta than IFLScience.)

The book’s humor really ties it together, often in surprisingly subtle ways. Each chapter has its own running joke, initially a throwaway line that grows into metaphors for everything the chapter discusses. It’s a great way to help the audience visualize without introducing too many new concepts at once. If there’s one thing cartoonists can teach science communicators, it’s the value of repetition.

I liked “We Have No Idea”. It could have been more daring, or more thorough, but it was still charming and honest and fun. If you’re looking for a Christmas present to explain physics to your relatives, you won’t go wrong with this book.

Why the Coupling Constants Aren’t Constant: Epistemology and Pragmatism

If you’ve heard a bit about physics, you might have heard that each of the fundamental forces (electromagnetism, the weak nuclear force, the strong nuclear force, and gravity) has a coupling constant, a number, handed down from nature itself, that determines how strong of a force it is. Maybe you’ve seen them in a table, like this:

If you’ve heard a bit more about physics, though, you’ll have heard that those coupling constants aren’t actually constant! Instead, they vary with energy. Maybe you’ve seen them plotted like this:

The usual way physicists explain this is in terms of quantum effects. We talk about “virtual particles”, and explain that any time particles and forces interact, these virtual particles can pop up, adding corrections that change with the energy of the interacting particles. The coupling constant includes all of these corrections, so it can’t be constant, it has to vary with energy.

Maybe you’re happy with this explanation. But maybe you object:

“Isn’t there still a constant, though? If you ignore all the virtual particles, and drop all the corrections, isn’t there some constant number you’re correcting? Some sort of `bare coupling constant’ you could put into a nice table for me?”

There are two reasons I can’t do that. One is an epistemological reason, that comes from what we can and cannot know. The other is practical: even if I knew the bare coupling, most of the time I wouldn’t want to use it.

The first thing to understand is that we can’t measure the bare coupling directly. When we measure the strength of forces, we’re always measuring the result of quantum corrections. We can’t “turn off” the virtual particles.

You could imagine measuring it indirectly, though. You’d measure the end result of all the corrections, then go back and calculate. That calculation would tell you how big the corrections were supposed to be, and you could subtract them off, solve the equation, and find the bare coupling.

And this would be a totally reasonable thing to do, except that when you go and try to calculate the quantum corrections, instead of something sensible, you get infinity.

We think that “infinity” is due to our ignorance: we know some of the quantum corrections, but not all of them, because we don’t have a final theory of nature. In order to calculate anything we need to hedge around that ignorance, with a trick called renormalization. I talk about that more in an older post. The key message to take away there is that in order to calculate anything we need to give up the hope of measuring certain bare constants, even “indirectly”. Once we fix a few constants that way, the rest of the theory gives reliable predictions.

So we can’t measure bare constants, and we can’t reason our way to them. We have to find the full coupling, with all the quantum corrections, and use that as our coupling constant.

Still, you might wonder, why does the coupling constant have to vary? Can’t I just pick one measurement, at one energy, and call that the constant?

This is where pragmatism comes in. You could fix your constant at some arbitrary energy, sure. But you’ll regret it.

In particle physics, we usually calculate in something called perturbation theory. Instead of calculating something exactly, we have to use approximations. We add up the approximations, order by order, expecting that each time the corrections will get smaller and smaller, so we get closer and closer to the truth.

And this works reasonably well if your coupling constant is small enough, provided it’s at the right energy.

If your coupling constant is at the wrong energy, then your quantum corrections will notice the difference. They won’t just be small numbers anymore. Instead, they end up containing logarithms of the ratio of energies. The more difference between your arbitrary energy scale and the correct one, the bigger these logarithms get.

This doesn’t make your calculation wrong, exactly. It makes your error estimate wrong. It means that your assumption that the next order is “small enough” isn’t actually true. You’d need to go to higher and higher orders to get a “good enough” answer, if you can get there at all.

Because of that, you don’t want to think about the coupling constants as actually constant. If we knew the final theory then maybe we’d know the true numbers, the ultimate bare coupling constants. But we still would want to use coupling constants that vary with energy for practical calculations. We’d still prefer the plot, and not just the table.

The Physics Isn’t New, We Are

Last week, I mentioned the announcement from the IceCube, Fermi-LAT, and MAGIC collaborations of high-energy neutrinos and gamma rays detected from the same source, the blazar TXS 0506+056. Blazars are sources of gamma rays, thought to be enormous spinning black holes that act like particle colliders vastly more powerful than the LHC. This one, near Orion’s elbow, is “aimed” roughly at Earth, allowing us to detect the light and particles it emits. On September 22, a neutrino with energy around 300 TeV was detected by IceCube (a kilometer-wide block of Antarctic ice stuffed with detectors), coming from the direction of TXS 0506+056. Soon after, the satellite Fermi-LAT and ground-based telescope MAGIC were able to confirm that the blazar TXS 0506+056 was flaring at the time. The IceCube team then looked back, and found more neutrinos coming from the same source in earlier years. There are still lingering questions (Why didn’t they see this kind of behavior from other, closer blazars?) but it’s still a nice development in the emerging field of “multi-messenger” astronomy.

It also got me thinking about a conversation I had a while back, before one of Perimeter’s Public Lectures. An elderly fellow was worried about the LHC. He wondered if putting all of that energy in the same place, again and again, might do something unprecedented: weaken the fabric of space and time, perhaps, until it breaks? He acknowledged this didn’t make physical sense, but what if we’re wrong about the physics? Do we really want to take that risk?

At the time, I made the same point that gets made to counter fears of the LHC creating a black hole: that the energy of the LHC is less than the energy of cosmic rays, particles from space that collide with our atmosphere on a regular basis. If there was any danger, it would have already happened. Now, knowing about blazars, I can make a similar point: there are “galactic colliders” with energies so much higher than any machine we can build that there’s no chance we could screw things up on that kind of scale: if we could, they already would have.

This connects to a broader point, about how to frame particle physics. Each time we build an experiment, we’re replicating something that’s happened before. Our technology simply isn’t powerful enough to do something truly unprecedented in the universe: we’re not even close! Instead, the point of an experiment is to reproduce something where we can see it. It’s not the physics itself, but our involvement in it, our understanding of it, that’s genuinely new.

The IceCube experiment itself is a great example of this: throughout Antarctica, neutrinos collide with ice. The only difference is that in IceCube’s ice, we can see them do it. More broadly, I have to wonder how much this is behind the “unreasonable effectiveness of mathematics”: if mathematics is just the most precise way humans have to communicate with each other, then of course it will be effective in physics, since the goal of physics is to communicate the nature of the world to humans!

There may well come a day when we’re really able to do something truly unprecedented, that has never been done before in the history of the universe. Until then, we’re playing catch-up, taking laws the universe has tested extensively and making them legible, getting humanity that much closer to understanding physics that, somewhere out there, already exists.

Why a New Particle Matters

A while back, when the MiniBoone experiment announced evidence for a sterile neutrino, I was excited. It’s still not clear whether they really found something, here’s an article laying out the current status. If they did, it would be a new particle beyond those predicted by the Standard Model, something like the neutrinos but which doesn’t interact with any of the fundamental forces except gravity.

At the time, someone asked me why this was so exciting. Does it solve the mystery of dark matter, or any other long-standing problems?

The sterile neutrino MiniBoone is suggesting isn’t, as far as I’m aware, a plausible candidate for dark matter. It doesn’t solve any long-standing problems (for example, it doesn’t explain why the other neutrinos are so much lighter than other particles). It would even introduce new problems of its own!

It still matters, though. One reason, which I’ve talked about before, is that each new type of particle implies a new law of nature, a basic truth about the universe that we didn’t know before. But there’s another reason why a new particle matters.

There’s a malaise in particle physics. For most of the twentieth century, theory and experiment were tightly linked. Unexpected experimental results would demand new theory, which would in turn suggest new experiments, driving knowledge forward. That mostly stopped with the Standard Model. There are a few lingering anomalies, like the phenomena we attribute to dark matter, that show the Standard Model can’t be the full story. But as long as every other experiment fits the Standard Model, we have no useful hints about where to go next. We’re just speculating, and too much of that warps the field.

Critics of the physics mainstream pick up on this, but I’m not optimistic about what I’ve seen of their solutions. Peter Woit has suggested that physics should emulate the culture of mathematics, caring more about rigor and being more careful to confirm things before speaking. The title of Sabine Hossenfelder’s “Lost in Math” might suggest the opposite, but I get the impression she’s arguing for something similar: that particle physicists have been using sloppy arguments and should clean up their act, taking foundational problems seriously and talking to philosophers to help clarify their ideas.

Rigor and clarity are worthwhile, but the problems they’ll solve aren’t the ones causing the malaise. If there are problems we can expect to solve just by thinking better, they’re problems that we found by thinking in the first place: quantum gravity theories that stop making sense at very high energies, paradoxical thought experiments with black holes. There, rigor and clarity can matter: to some extent they’re already there, but I can appreciate the argument that it’s not yet nearly enough.

What rigor and clarity won’t do is make physics feel (and function) like it did in the twentieth century. For that, we need new evidence: experiments that disobey the Standard Model, and do it in a clear enough way that we can’t just chalk it up to predictable errors. We need a new particle, or something like it. Without that, our theories are most likely underdetermined by the data, and anything we propose is going to be subjective. Our subjective judgements may get better, we may get rid of the worst-justified biases, but at the end of the day we still won’t have enough information to actually make durable progress.

That’s not a popular message, in part, because it’s not something we can control. There’s a degree of helplessness in realizing that if nature doesn’t throw us a bone then we’ll probably just keep going in circles forever. It’s not the kind of thing that lends itself to a pithy blog post.

If there’s something we can do, it’s to keep our eyes as open as possible, to make sure we don’t miss nature’s next hint. It’s why people are getting excited about low-energy experiments, about precision calculations, about LIGO. Even this seemingly clickbaity proposal that dark matter killed the dinosaurs is motivated by the same sort of logic: if the only evidence for dark matter we have is gravitational, what can gravitational evidence tell us about what it’s made of? In each case, we’re trying to widen our net, to see new phenomena we might have missed.

I suspect that’s why this reviewer was disappointed that Hossenfelder’s book lacked a vision for the future. It’s not that the book lacked any proposals whatsoever. But it lacked this kind of proposal, of a new place to look, where new evidence, and maybe a new particle, might be found. Without that we can still improve things, we can still make progress on deep fundamental mathematical questions, we can kill off the stupidest of the stupid arguments. But the malaise won’t lift, we won’t get back to the health of twentieth century physics. For that, we need to see something new.

Amplitudes 2018

This week, I’m at Amplitudes, my field’s big yearly conference. The conference is at SLAC National Accelerator Laboratory this year, a familiar and lovely place.

Welcome to the Guest House California

It’s been a packed conference, with a lot of interesting talks. Recording and slides of most of them should be up at this point, for those following at home. I’ll comment on a few that caught my attention, I might do a more in-depth post later.

The first morning was dedicated to gravitational waves. At the QCD Meets Gravity conference last December I noted that amplitudes folks were very eager to do something relevant to LIGO, but that it was still a bit unclear how we could contribute (aside from Pierpaolo Mastrolia, who had already figured it out). The following six months appear to have cleared things up considerably, and Clifford Cheung and Donal O’Connel’s talks laid out quite concrete directions for this kind of research.

I’d seen Erik Panzer talk about the Hepp bound two weeks ago at Les Houches, but that was for a much more mathematically-inclined audience. It’s been interesting seeing people here start to see the implications: a simple method to classify and estimate (within 1%!) Feynman integrals could be a real game-changer.

Brenda Penante’s talk made me rethink a slogan I like to quote, that N=4 super Yang-Mills is the “most transcendental” part of QCD. While this is true in some cases, in many ways it’s actually least true for amplitudes, with quite a few counterexamples. For other quantities (like the form factors that were the subject of her talk) it’s true more often, and it’s still unclear when we should expect it to hold, or why.

Nima Arkani-Hamed has a reputation for talks that end up much longer than scheduled. Lately, it seems to be due to the sheer number of projects he’s working on. He had to rush at the end of his talk, which would have been about cosmological polytopes. I’ll have to ask his collaborator Paolo Benincasa for an update when I get back to Copenhagen.

Tuesday afternoon was a series of talks on the “NNLO frontier”, two-loop calculations that form the state of the art for realistic collider physics predictions. These talks brought home to me that the LHC really does need two-loop precision, and that the methods to get it are still pretty cumbersome. For those of us off in the airy land of six-loop N=4 super Yang-Mills, this is the challenge: can we make what these people do simpler?

Wednesday cleared up a few things for me, from what kinds of things you can write down in “fishnet theory” to how broad Ashoke Sen’s soft theorem is, to how fast John Joseph Carrasco could show his villanelle slide. It also gave me a clearer idea of just what simplifications are available for pushing to higher loops in supergravity.

Wednesday was also the poster session. It keeps being amazing how fast the field is growing, the sheer number of new faces was quite inspiring. One of those new faces pointed me to a paper I had missed, suggesting that elliptic integrals could end up trickier than most of us had thought.

Thursday featured two talks by people who work on the Conformal Bootstrap, one of our subfield’s closest relatives. (We’re both “bootstrappers” in some sense.) The talks were interesting, but there wasn’t a lot of engagement from the audience, so if the intent was to make a bridge between the subfields I’m not sure it panned out. Overall, I think we’re mostly just united by how we feel about Simon Caron-Huot, who David Simmons-Duffin described as “awesome and mysterious”. We also had an update on attempts to extend the Pentagon OPE to ABJM, a three-dimensional analogue of N=4 super Yang-Mills.

I’m looking forward to Friday’s talks, promising elliptic functions among other interesting problems.

If you’ve ever heard someone list problems in academia, citation-counting is usually near the top. Hiring and tenure committees want easy numbers to judge applicants with: number of papers, number of citations, or related statistics like the h-index. Unfortunately, these metrics can be gamed, leading to a host of bad practices that get blamed for pretty much everything that goes wrong in science. In physics, it’s not even clear that these statistics tell us anything: papers in our field have been including more citations over time, and for thousand-person experimental collaborations the number of citations and papers don’t really reflect any one person’s contribution.

It’s pretty easy to find people complaining about this. It’s much rarer to find a proposed solution.

That’s why I quite enjoyed Alessandro Strumia and Riccardo Torre’s paper last week, on Biblioranking fundamental physics.

Some of their suggestions are quite straightforward. With the number of citations per paper increasing, it makes sense to divide each paper by the number of citations it contains: it means more to get cited by a paper with ten citations than by a paper with one hundred. Similarly, you could divide credit for a paper among its authors, rather than giving each author full credit.

Some are more elaborate. They suggest using a variant of Google’s PageRank algorithm to rank papers and authors. Essentially, the algorithm imagines someone wandering from paper to paper and tries to figure out which papers are more central to the network. This is apparently an old idea, but by combining it with their normalization by number of citations they eke a bit more mileage from it. (I also found their treatment a bit clearer than the older papers they cite. There are a few more elaborate setups in the literature as well, but they seem to have a lot of free parameters so Strumia and Torre’s setup looks preferable on that front.)

One final problem they consider is that of self-citations, and citation cliques. In principle, you could boost your citation count by citing yourself. While that’s easy to correct for, you could also be one of a small number of authors who cite each other a lot. To keep the system from being gamed in this way, they propose a notion of a “CitationCoin” that counts (normalized) citations received minus (normalized) citations given. The idea is that, just as you can’t make anyone richer just by passing money between your friends without doing anything with it, so a small community can’t earn “CitationCoins” without getting the wider field interested.

There are still likely problems with these ideas. Dividing each paper by its number of authors seems like overkill: a thousand-person paper is not typically going to get a thousand times as many citations. I also don’t know whether there are ways to game this system: since the metrics are based in part on citations given, not just citations received, I worry there are situations where it would be to someone’s advantage to cite others less. I think they manage to avoid this by normalizing by number of citations given, and they emphasize that PageRank itself is estimating something we directly care about: how often people read a paper. Still, it would be good to see more rigorous work probing the system for weaknesses.

In addition to the proposed metrics, Strumia and Torre’s paper is full of interesting statistics about the arXiv and InSpire databases, both using more traditional metrics and their new ones. Whether or not the methods they propose work out, the paper is definitely worth a look.

Path Integrals and Loop Integrals: Different Things!

When talking science, we need to be careful with our words. It’s easy for people to see a familiar word and assume something totally different from what we intend. And if we use the same word twice, for two different things…

I’ve noticed this problem with the word “integral”. When physicists talk about particle physics, there are two kinds of integrals we mention: path integrals, and loop integrals. I’ve seen plenty of people get confused, and assume that these two are the same thing. They’re not, and it’s worth spending some time explaining the difference.

Let’s start with path integrals (also referred to as functional integrals, or Feynman integrals). Feynman promoted a picture of quantum mechanics in which a particle travels along many different paths, from point A to point B.

You’ve probably seen a picture like this. Classically, a particle would just take one path, the shortest path, from A to B. In quantum mechanics, you have to add up all possible paths. Most longer paths cancel, so on average the short, classical path is the most important one, but the others do contribute, and have observable, quantum effects. The sum over all paths is what we call a path integral.

It’s easy enough to draw this picture for a single particle. When we do particle physics, though, we aren’t usually interested in just one particle: we want to look at a bunch of different quantum fields, and figure out how they will interact.

We still use a path integral to do that, but it doesn’t look like a bunch of lines from point A to B, and there isn’t a convenient image I can steal from Wikipedia for it. The quantum field theory path integral adds up, not all the paths a particle can travel, but all the ways a set of quantum fields can interact.

How do we actually calculate that?

One way is with Feynman diagrams, and (often, but not always) loop integrals.

I’ve talked about Feynman diagrams before. Each one is a picture of one possible way that particles can travel, or that quantum fields can interact. In some (loose) sense, each one is a single path in the path integral.

Each diagram serves as instructions for a calculation. We take information about the particles, their momenta and energy, and end up with a number. To calculate a path integral exactly, we’d have to add up all the diagrams we could possibly draw, to get a sum over all possible paths.

(There are ways to avoid this in special cases, which I’m not going to go into here.)

Sometimes, getting a number out of a diagram is fairly simple. If the diagram has no closed loops in it (if it’s what we call a tree diagram) then knowing the properties of the in-coming and out-going particles is enough to know the rest. If there are loops, though, there’s uncertainty: you have to add up every possible momentum of the particles in the loops. You do that with a different integral, and that’s the one that we sometimes refer to as a loop integral. (Perhaps confusingly, these are also often called Feynman integrals: Feynman did a lot of stuff!)

$\frac{i^{a+l(1-d/2)}\pi^{ld/2}}{\prod_i \Gamma(a_i)}\int_0^\infty...\int_0^\infty \prod_i\alpha_i^{a_i-1}U^{-d/2}e^{iF/U-i\sum m_i^2\alpha_i}d\alpha_1...d\alpha_n$

Loop integrals can be pretty complicated, but at heart they’re the same sort of thing you might have seen in a calculus class. Mathematicians are pretty comfortable with them, and they give rise to numbers that mathematicians find very interesting.

Path integrals are very different. In some sense, they’re an “integral over integrals”, adding up every loop integral you could write down. Mathematicians can define path integrals in special cases, but it’s still not clear that the general case, the overall path integral picture we use, actually makes rigorous mathematical sense.

So if you see physicists talking about integrals, it’s worth taking a moment to figure out which one we mean. Path integrals and loop integrals are both important, but they’re very, very different things.