Thursday, 16 January 2020

How to make progress in High Energy Physics

How to make progress in High Energy Physics Before I start, just following from my previous post about B-mesons, today I saw a CERN press release about lepton universality in B-baryons (i.e. particles made of three quarks, at least one of which is a bottom, rather than B-mesons, which have two quarks, at least one of which is a bottom). It seems there is a \( 1 \sigma \) deviation in $$ R_{pK}^{-1} \equiv \frac{\mathrm{BR} (\Lambda_b^0 \rightarrow p K^- e^+ e^-)}{\mathrm{BR} (\Lambda_b^0 \rightarrow p K^- J/\psi(\rightarrow e^+ e^-))} \times \frac{\mathrm{BR} (\Lambda_b^0 \rightarrow p K^- J/\psi(\rightarrow \mu^+ \mu^-))}{\mathrm{BR} (\Lambda_b^0 \rightarrow p K^- \mu^+ \mu^-)} $$ While, by itself, it is amazing that this gets a press release heralding a "crack in the Standard Model", it does add some small evidence to the picture of deviations from Standard Model predictions; no doubt the interpretation in terms of a global fit with other observables will appear soon on the arXiv. So it's a positive way to start this entry.

Polemics on foundations of HEP

Recently an article appeared by S. Hossenfelder that again makes the claim that "fundamental physics" is stuck, has failed etc, that theorists are pursuing dead-end theories and "do not think about which hypotheses are promising" and "theoretical physicists have developed a habit of putting forward entirely baseless speculations." It is fairly common and depressing to see this message echoed in a public space. However what riled me enough to write was the article on Not Even Wrong discussing it, in which surprise is expressed that people at elite institutions would still teach courses in Beyond the Standard Model Physics and Supersymmetry. This feels to me like an inadvertent personal attack, since I happen to teach courses on BSM physics and SUSY at an elite French institution (Ecole Polytechnique) ... hence this post.

Physicists are not sheep

Firstly though I'd like to address the idea that physicists are not aware of the state of their field. While "maverick outsiders" might like to believe that HEP theorists live in a bubble, just following what they are told to work on, it upsets me that this message has cut through enough that a lot of students now feel that it is true, that if they do not work on what they perceive to be "hot topics" then they will be censured. The truth is that now, more than perhaps at any time since I entered the field, there is a lack of really "hot topics." In previous decades, there were often papers that would appear with a new idea that would be immediately jumped on by tens or hundreds of people, leading to a large number of more-or-less fast follow-up papers (someone once categorised string theorists as monkeys running from tree to tree eating only the low-hanging fruit). This seems to me to be much less prevalent now. People are really stepping back and thinking about what they do. They are aware that there is a lack of clear evidence for what form new physics should take, and that the previously very popular idea of naturalness is probably not a reliable guide. Some people are embracing new directions in dark matter searches, others are trying to re-interpret experimental data in terms of an effective field theory extension of the Standard Model, others are seeing what we can learn from gravitational waves, still others are looking at developing new searches for axions, some people are looking instead at fundamental Quantum Field Theory problems, cosmology has made huge progress, etc etc (see also this thread by Dan Green). There is a huge diversity of ideas in the field and that is actually very healthy. There is also a very healthy amount of skepticism.

On the other hand, as I mentioned in my previous post, there are some tentative pieces of evidence pointing to new physics at accessible scales; and whatever explains dark matter, it should be possible to probe it with some form of experiment or observation. This is the reason for continued optimism in the field of real breakthroughs. We could be on the verge of overturning the status quo, via the (apparently outdated) method of doing experiments, and then we will race to understand the results and interpret them in terms of our favourite theories -- or maybe genuinely new ones. Of course, maybe these are mirages; as physicists we will continue to look for new and creative ways to search for new phenomena, even if we do not have a new high energy collider -- but if we don't build a new collider we will never know what we might find.

What courses should you take

Coming now to the idea of what students entering the field should learn, in the current negative climate it needs repeating that the Standard Model is incomplete. I'm not just talking about a lack of quantum gravity, but there is a laundry list of problems that I repeat to my students at the beginning of the course:
  1. Quantum gravity.
  2. Dark matter, or something that explains rotation curves, the CMB, etc.
  3. Dark energy -- no, it hasn't been ruled out by one paper on supernovae. It was awarded the Nobel Prize because people already expected it from other observations.
  4. Inflation, or something else that solves the same problems.
  5. The strong CP problem. We have phases in the quark Yukawas, so we should have a neutron electric dipole moment \(10^{10} \) times greater than we observe. Most people believe this should be solved by an axion -- which might also be dark matter -- hence a lot of effort to find it, and ADMX (among other experiments) might be getting close.
  6. Baryogenesis. The Standard Model Higgs is too heavy to have electroweak baryogenesis. There is apparently not enough CP violation in the Standard Model either.
  7. Neutrino masses. We can't write them into the Standard Model because we don't even know if neutrinos are Majorana or Dirac! Maybe a heavy right-handed neutrino can give us Baryogenesis through leptogenesis. There is a huge amount going on in neutrino physics at the moment, too ...

Nearly all of these topics are not generally covered in a standard set of graduate courses (at least here). I try to present the evidence and some possible solutions. So the first time a lot of students encounter these issues is through popular press articles, and oblique references in "standard" courses. And if we are going to make progress on solving some of these fundamental issues, should students not have some idea on what attempts have been made to solve them?

Turning now to supersymmetry, I would not recommend that a beginning student in particle phenomenology make it the sole focus of their work (unless they really have a good motivation to do so). But there are many reasons to study it still:

  1. It is hugely important in formal applications -- to give us a handle on strongly coupled theories, allowing us to compute things we could never do in non-SUSY theories, as toy models, \( N=4 \) SYM being the "simplest field theory" (as Arkani-Hamed likes to reiterate) etc etc.
  2. It seems to be necessary for the consistency of string theory. I personally prefer string theory as a candidate framework for quantum gravity; if you want to study it, you need to study SUSY.
  3. A lot of the difficulty with the formalism for beginning students is just understanding two-component spinors -- these are actually very useful tools if you want to study amplitudes in general.
  4. It allows us to actually address the Hierarchy problem, and related to this, the idea of the vacuum energy of the theory being related to a cosmological constant. This is a subtle (and maybe heated) discussion for another time.
  5. The gauge couplings apparently unify in the simplest SUSY extensions of the Standard Model. If this is just a coincidence then I feel that nature is playing a cruel joke on us.
  6. The Standard Model appears to be at best metastable (there is some dispute about this). It has been further suggested (e.g. here) that black holes might seed the vacuum decays, so that if it is not absolutely stable then it should decay much quicker than we would otherwise think; and in any case the standard calculation has to assume that there are no quantum gravity contributions (giving higher-order operators). New physics at an intermediate scale (below \( \sim 10^{11} \) GeV) such as supersymmetry would then be necessary to stabilise the vacuum.
  7. It genuinely could still be found at accessible energies; the LHC is actually very poor at finding particles that don't couple to the strong force, and new electroweak states could easily be lurking in plain sight ...
  8. ... related to this, it's just about the only "phenomenological framework" for new physics that addresses lots of different problems with the Standard Model.

Of course, nowadays as a community we are trying to hedge our bets: there is much more ambivalence about what theories might be found just around the corner, hence my own work on generic pheneomenology, and a lot of interest in the Standard Model EFT.

How we should make progress

Finally we get to the topic of the post. In the original article that I linked to above, Hossenfelder does make (as she has made elsewhere) the positive suggestion that physicists should talk to philosophers. [ In France, this is amusing, because there is a fantastic tradition of famous philosophers, and every schoolchild has to study philosophy up to the age of 18 ] It is good to make suggestions. In the article though is the idea that people cannot recognise promising new ideas amidst a sea of "bad" ones, so people are either following old dead ends or endlessly making ridiculous suggestions. I admit that, superficially, this is the impression people could have got once upon a time, but I would argue is not the state of the field now. I disagree that the problem is a fundamental one about how people think, or that there is a system censuring of "good" radical ideas. I don't think there is only one way to make progress: if I had a suggestion how the scientific creative process should work I would be applying it like crazy before advertising the benefits publically! And of censureship of "good" ideas, there are a lot of people willing to take a risk on new concepts. Research is hard, and creativity is not something that is easily taught. But I am constantly amazed by the creativity and ingenuity of my peers, the diversity of their ideas, and it is heartbreaking to see their effort denigrated in popular articles.

Indeed, repeatedly making the claim in public that one group of scientists are dishonest (or kidding themselves) about progress, that the field has failed etc, helps no-one. It deeply worries me to read that Dominic Cummings has Not Even Wrong on his blog roll; and I have already seen that people in other fields often hold very wrong opinions about the state of fundamental physics due to this filtering through (most people only see wildly speculative and hagiographic articles on one side and hugely negative pessimism on the other). It has been an issue when deciding about grant funding for at least a decade already. And it also filters through to students when they are deciding what to do, who, as I pointed out above, usually haven't really seen enough about the fundamentals of the field before they have to make a decision on what they want to study.

Finally, coming back to the suggestion that physicists do not think about what they are doing or why, there are two very important times when we emphatically do do this: when we are teaching, and when we are writing grant proposals. The preparation of both of these things can be hard work, but they are rewarding, and more reasons that I have faith in my fellow physicists ability to genuinely try to challenge the big problems in our field.


  1. It's not a personal attack, but people really should take seriously the argument that the uniformly negative experiment/theory results of the last thirty years concerning SUSY extensions of the SM imply that it is no longer a good idea to train graduate students in this material. Concerning your 8 arguments for continuing to do so:

    1. Sure, people working on non-perturbative QFTs and on topological quantum field theories should know about N=4 SYM. But the motivations there are very different, not possible BSM physics in the real world. The Arkani-Hamed "simplest QFT" or "harmonic oscillator of the 21st century" claims are the kind of hype students should be protected from, not fed.

    2. At this point, the better argument is that the connection between SUSY and string theory has become a reason not to learn about SUSY, because of the failure of string theory unification.

    3. Agreed that understanding spinors is tricky, trying to do so in the much more complicated context of SUSY theories just makes this harder.

    4. SUSY in no way helps with the vacuum energy problem, arguably makes it much worse (SUSY breaking makes the problem much more well-defined and robust).

    5. A single numerical coincidence is not much to go on.

    6/7/8. New BSM physics would be great, the problem is that thirty years of negative results have moved the SUSY paradigm from one with a motivation you could argue for to an unmotivated one.

  2. This comment has been removed by the author.

  3. The fundamental issue here is that the motivations for teaching supersymmetry listed here are poor, and there are much better motivations for teaching supersymmetry to graduate students. Supersymmetry is an extremely useful tool for studying the dynamics in strongly coupled quantum field theories and quantum gravity, and arguably this role has become the main focus of theoretical physicists studying supersymmetry in the past few years, over beyond standard model phenomenology. Furthermore, through theories such as Parisi's supersymmetric theory of stochastic dynamics, supersymmetry has become a useful tool to analyse stochastic partial differential equations with applications to other areas of physics such as classical statistical mechanics, optics, and condensed matter physics (the supersymmetric WKB approximation is a good example of that), or even in biology or finance. However, the null results in the past 40 years have shown that its relevance to beyond standard model physics is at best extremely diminished.

  4. Hi Peter, thanks for your comments.

    1. I don't understand what is wrong with the claim that N=4 SYM is/could be the simplest field theory. It is *the* toy model used by legions of amplitude theorists precisely because it is simple.

    2. As with point 1, it's not necessarily anything to do with low energy SUSY pheno. You can use string theory for a number of things. (And of course it may yet have something to do with the real world. We won't agree on that though!)

    3. I disagree that SUSY complicates things. At the beginning the formalism appears daunting, and then it all moves out of the way and becomes simple. SUSY gives a very concrete application, and it is nearly always better to learn things through examples.

    4. I do not know why making the problem more well-defined and robust is worse.

    5. I would not bet anything significant on a single coincidence either. But it would be strange to ignore it! When you have few clues, you have to make the most of the ones you have! And it at least forms part of a consistent picture.

    6/7/8. As I said above, I don't generally recommend students entering the field to study exclusively SUSY. But it is not unmotivated. Even low energy SUSY as physics Beyond The Standard Model (as opposed to split, high-scale, etc SUSY) does still have many motivations, particularly as a framework. People should keep looking at/for it and be aware of it; most people know that now we also need to be thinking of alternatives. The problem is that people have been trying to find something better for a long time already, so this either tells us something by itself or represents a real opportunity.

  5. Hi Mark,
    A couple quick responses:
    1. I don't want to argue against studying N=4 SYM as a very non-trivial example of a strongly interacting QFT, unusually tractable because of its symmetries. But the full structure of the model itself is rather complicated, with a lot of degrees of freedom conspiring in a tricky way. To me, describing this situation as "simple" is misleading.

    4. It's the difference between your theory coming with a not very solid argument for a deadly problem (vacuum energy off by a large number of orders of magnitude) and coming with a very solid argument for the deadly problem. The second alternative is more deadly.

  6. "... supersymmetry ... It seems to be necessary for the consistency of string theory ... as a candidate framework for quantum gravity; if you want to study it, you need to study SUSY." It seems to me that the preceding judgment is correct for string theory with the infinite nature hypothesis. My guess is that string theory with the finite nature hypothesis implies Milgrom's MOND and the Riofrio-Sanejouand cosmological model. The Riofrio model might be wrong, but there is no way that MOND is wrong. Consider the Milgrom Denial Hypothesis: The main problem with string theory is that string theorists fail to realize that Milgrom is the Kepler of contemporary cosmology. My experience with attempting to convince string theorists that MOND is empirically valid indicates that these string theorists indicate that MOND is a mistake based on data dredging. These string theorists are 100% wrong. Google "kroupa milgrom" and "scarpa milgrom".

  7. Supersymmetry is valid theory, for example many solid-state quasiparticles, gluebals or Hungarian boson fulfils phenomenology of supersymmetric particles quite well. I even consider indicia of Higgs pentuplet which evaded attention in LHC experiments, because it emerges in dilepton decay channels only. It's just important to realize, SuSy is extradimensional theory, it doesn't apply to free particles but only surface phenomena / low-dimensional geometric arrangement of normal particles, where Casimir vacuum and/or virtual quarks may apply. This also renders SuSy as a way less fundamental, significant and widespread, than its proponents consider. And of course they also missed mass/energy density scale at which SuSy manifest itself (this aspect has SuSy common with string theory and its extradimensions). In bizarre turn of events the SuSy may apply just to phenomena which mainstream physics learned to ignore most: to antigravity / overunity effects and scalar wave physics of Nicola Tesla.

    1. What dark matter, Yukawa and Casimir forces have in common? Being hyperdimensional forces), their force constant changes with distance way more pronouncedly, that gravitational or let say Coulomb force (this aspect they have common with nuclear force). But they're also directional (at the end of elongated shapes) they're getting stronger, which is why no single boson can be attributed to them (they're unparticles). These forces thus glue massive bodies at distance like filaments of glue, this aspect they have common with gluons (which also come with multiple varieties). Under certain geometric arrangement (during decay of dumbbell shaped nuclei and atoms) these forces can still manifest itself with multiple distinct peaks on boson decay spectrum - this aspect they have common with SUSY particles - such a bosons are therefore superpartners of interacting particles at the same moment. The phenomenology of these forces is thus quite colourful and in many aspects it resembles complex behaviour of quasiparticles of solid state physics.

  8. /* there is no way that MOND is wrong */

    MOND is violated by every dark matter filament, by galaxies with anomalously high or low dark matter content or by Bullet cluster or by dark matter fragments recently observed. But it doesn't mean that this theory is wrong in similar way, like parabola is wrong model of waterfall because real waterfall exhibit deviations from it. It's just approximate model of warm cold dark matter, whereas dark matter also comes with cold and hot variants. It applies to certain distance scales and certain (albeit quite widespread) arrangement of massive bodies.

    Similarly to SuSy phenomenology one cannot expect that description of hyperdimensional effects like dark matter and MOND will remain universally valid with the same precision, like QED for example. We should perceive these theories as a certain low-dimensional (holographic if you wish) projection of hyperdimensional reality.

    Ironically just the theorists of SuSy accustomed to deterministic behavior of low-dimensional theories are the greatest obstacle of understanding of SuSy phenomenology. And I'm not even talking about proponents of quantum gravity like Hossenfelder and Woit.

    1. "MOND is violated by every dark matter filament ..." Let us assume that dark matter filaments are quite ubiquitous in our universe. Does that assumption contradict MOND? I say no. MOND from its inception has been data-driven. I think that, to some extent, MIlgrom has spread confusion concerning MOND. He seems to think that Bekenstein's TeVeS might be on the right track — it seems to me that, from the stringy viewpoint, TeVeS is completely wrong. I think that MOND has many empirical successes and there are 2 basic alternatives: (1) Something is slightly wrong with Newtonian-Einstein gravitational theory. (2) After quantum averaging, Einstein's field equations are 100% correct, but some of the dark matter particles have MOND-compatible properties. If (2) is correct, then my guess is that some dark matter particles are MOND-chameleon particles that have variable effective mass depending upon nearby gravitational acceleration. I think that string theorists might be able to provide a model for MOND by assuming that gravitons have one or more D-brane charges and that gravitinos have one or more D-brane charges and somehow these hypothetical D-brane charges "conspire" to simulate MOND.

    2. MOND/MOD, TeVeS/STVG, QI/MiHsC are extension of general relativity, their dark matter thus can have shape of sphere or disk at best case. Dark matter filaments are thus based on different mechanism, which can be characterized by particle models of dark matter and/or deDuillier/LeSage gravity theory. In particular so called lanterns formed along black hole jets can be domain of supersymmetric particles.

    3. The question is: What is relativistic MOND? If dark-matter-compensation-constant = 0, then my speculations about the foundations of physics are wrong, but (non-relativistic) MOND has many empirical successes. Google "tully-fisher mond".

  9. Just to amplify the lead-in remark:
    While the flavour anomaly now reported in Lambda_b decays is only at 1 sigma, it points in the same direction as that seen in other channels, with joint significance around 4 sigma: see here.