Of Particular Significance

A Second Higgs Particle?

POSTED BY Matt Strassler

POSTED BY Matt Strassler

ON 07/02/2013

It is well-known in science that if the title of a post or paper begins with a question, the answer is always “NO”, or at best, “probably not”.  Today we’re working with “very probably not”.

Yes, we’ve found one type of Higgs particle, but there might be two, three, or even more types of Higgs particles in nature.  Such particles might well be discovered eventually at the Large Hadron Collider [LHC] or at future experiments hardly dreamt of today.  But at present, there’s no evidence yet for a second Higgs particle.  And all the hullabaloo we’re hearing right now is about old news, reprocessed into new news, which actually wasn’t news anyway and really isn’t now either.

The “News”

Here’s what all the fuss is about: it’s about old data, taken in 2011 and 2012 at the CMS experiment at the LHC, in the search for a Higgs particle decaying to two particles of light (“photons”).  This data was already fully processed by March 2013.  Here’s what CMS found, along with its competitor, ATLAS.

Data from CMS (left) and ATLAS (right) looking at all proton-proton collisions that each observed in which two isolated photons were produced.  What is plotted is the number of events versus the mass of any particle that could have produced those two photons.  A real particle will show up as a bump.  The Higgs particle at around 125 GeV is real; it appears in both ATLAS and CMS.  The bump at 136 GeV in CMS is not reproduced at ATLAS; there is therefore no evidence that it is a real particle, and it is likely to be just a statistical fluke.
From March 2013: Data from proton-proton collisions at CMS (left) and at ATLAS (right). The number of collisions that produced two energetic photons is on the vertical axis; the mass of any particle that could have produced those two photons, on the horizontal axis. A real particle will show up as a bump. The Higgs particle at around 125 GeV is real; it appears in both ATLAS and CMS. The bump at 136 GeV in CMS is not reproduced at ATLAS; consequently there is no evidence that it is due to a real particle, and it is likely to be just a statistical fluke.  (The lower ATLAS plot shows the data in the upper plot minus the red dotted curve in the upper plot.)

The evidence for the Higgs particle discovered in 2012, with a mass of 125-126 GeV/c², is circled in blue; it appears in both experiments.  I have also circled, in green, a region around 136 GeV/c².  What do you see?  You see a bump in the CMS data.  You don’t see a bump in the ATLAS data.

That bump at CMS is what all the latest excitement is about.  And it was sitting in the data back in March; why suddenly there’s all this talk in July is (almost) beyond me.

Let’s start with the basic question: what could this bump in the CMS data be?

  • Could it be a new particle? Maybe a 2nd type of Higgs particle, maybe something else?  Well if it is, then (just like the Higgs at 125 GeV/c²) it has to show up both in ATLAS and in CMS.  And right now, this thing doesn’t.
  • Could it be a statistical fluke, a bit large but not really that unusual?  Well, if it is, it is likely* to show up only at CMS and not in ATLAS, or vice versa.  And that, of course, is exactly what we see in the data.

[*Why? The two experiments take entirely independent data (the proton-proton collisions that ATLAS observes occur inside ATLAS, and the proton-proton collisions that CMS observes occur inside CMS), so if an unusual statistical accident occurred in one experiment’s data, it would be doubly unusual if the same statistical accident occurred in the other’s data… thus if it really is a fluke, we would expect to see it in only one experiment, not in both.]

Is Such a Big Fluke Unusual?

How big a statistical fluke is it really? Well, this depends how you analyze the data.  And all this week’s excitement comes from a new analysis of the same data.    Let me say that again: There’s no new data, just a different way of looking at it — which among other things means that the new analysis cannot have been done blind. When you make adjustments to the way you look at data, the significance of various bumps will go up and down a bit — and you shouldn’t necessarily take those changes too seriously.  Let us not forget that the significance of the excess that eventually became the Higgs particle at 125 GeV/c² went up and down too, over time, before firming up into something convincing.

In any case, the most optimistic analysis is that this bump at 135 GeV exceeds the expectation (the red curve on the plot) by 2.9 standard deviations (also called 2.9 “sigma”).  That’s almost 1 standard deviation higher than was quoted in March.  [Note also that this number does not include the look-elsewhere effect, so its significance is lower than it first sounds.] Why did the significance go up in the new paper?  Because after reanalysis (specifically, after including the known properties of the Higgs found at 125 GeV/c²), the red curve on the CMS plot shifted down slightly.  So that made the bump at 136  seem a bit bigger.

Particle physicists consider 3 standard deviations to be weak evidence for something.  So at best, CMS has something that is, let’s say, slightly below weak evidence.  An excess of this magnitude is a bit unusual, but we’ve certainly seen such things before in other plots from the LHC and other experiments, only to see them go away with time as more data was accrued.

But none of this matters, because ATLAS actually has no bump at 136 GeV.  Even if ATLAS’s red curve went down in the same way when reanalyzed, they still wouldn’t have a bump.  Since they had no excess at all in the March analysis, they aren’t going to get more than 1 standard deviation excess no matter what they do to massage the data.  So when you combine ATLAS and CMS’s independent information, it is obvious that the total statistical significance of what we’re seeing is well below 3 standard deviations — probably below 2.  In short, taking all of the LHC’s data, and not cherry-picking (i.e., selecting CMS because it looks interesting to you and ignoring ATLAS because it looks less interesting to you is cherry-picking — we’ve seen that kind of thing before in the media), there’s nothing here that you can call “evidence”, either technically or colloquially.

And this is presumably why, instead of suggesting they might have evidence for a new particle, CMS makes no claim at all in their paper.  They simply point out there’s a bump in their data with a significance approaching 3 standard deviations, and they make no claim that it represents something real.  In other words, they’re being responsible scientists. Other scientists and the media are less responsible.

A Final Note on Multiple Higgs Particles

Still, since I’ve been asked by reporters whether we’d expect to see a second Higgs particle, let me make, independent of this data, the following comments:

  1. No one will be surprised if there is a second type of Higgs particle in nature.  A very wide variety of speculative ideas that go beyond the “Standard Model” (the equations that describe the known particles and forces) predict the existence of more than one Higgs particle.  So don’t be surprised if there are more of them; I won’t be.  I just don’t see any evidence yet in the data.
  2. However, everyone will be surprised if there are two Higgs particles that both are similar to a Standard Model Higgs particle, the simplest possible type of Higgs.  In fact, the more Standard-Model-like the first Higgs particle seems to be, the more likely it is that the second Higgs particle is rarely produced.  And since the Higgs particle discovered last year appears to be Standard-Model-like, we would not generally expect a second Higgs to be visible on the CMS and ATLAS plot above; the bump should be much smaller.  But this expectation is not airtight; you can easily invent speculative theories in which the second Higgs decays so often to two photons that its bump is visible on this plot.
  3. Don’t forget, however, that a bump in the two-photon data from ATLAS and CMS wouldn’t necessarily imply a new Higgs particle anyway.  There are other types of particles that could give bumps in this data!

Share via:

Twitter
Facebook
LinkedIn
Reddit

55 Responses

  1. Hey! This post couldn’t be written any better! Reading through
    this post reminds me of my old room mate! He
    always kept chatting about this. I will forward this post
    to him. Pretty sure he will have a good read. Many thanks for sharing!

  2. I absolutely love your blog and find almost
    all of your post’s to be just what I’m looking for. Do you offer guest writers to
    write content for you personally? I wouldn’t mind creating a post or elaborating on a few of the
    subjects you write in relation to here. Again, awesome web
    site!

  3. Can particles exist at different time frames (different time intervals between states? So the graviton could be oscillating at a very low time interval since the big bang.

    Hence we live at the speed of light while more fundamental particles live at much faster speeds.

    Could a variable time be a way to factor in infinity to the math?

  4. correction: Is there any important news from Icecube experiment at the ICRC 2013 conference taking place now in Buenos Aires?

  5. isnt it strange that the 136 GeV CMS peak was there (tough with smaller significance) even when they presented around 5 @ 7 TeV + 5 @ 8 TeV inverse fb of data. Now after adding around 15 inverse fb of data at 8 TeV, the peak consistently remains there, even increasing its significance. ¿?

    1. No, it’s not strange. The probability of this happening is not small. Coincidences happen every day. That’s why physicists calculate the statistical significance of things like this — to determine whether this is a big coincidence or a small coincidence. Given that ATLAS did not see anything similar, the probability that this is a coincidence is high.

  6. “No one will be surprised if there is a second type of Higgs particle in nature. A very wide variety of speculative ideas that go beyond the “Standard Model” (the equations that describe the known particles and forces) predict the existence of more than one Higgs particle.”

    I would suggest that caveat that lots of beyond the Standard Model physics has more than one kind of Higgs particles (most notably all supersymmetry theories), but I can’t think of any that have one extra Higgs type particle and nothing else at all. Models that give you an extra Higgs type particle generally also give you many other undiscovered particles all of which have not yet been observed. So one faces the challenge if one sees just one extra bump in data of figuring out why all the other expected bumps that should go with it are missing (generally, they are assumed to be exiled to energy scales we can’t measure).

    1. Hi Ohwilleke, You wrote: Models that give you an extra Higgs type particle generally also give you many other undiscovered particles all of which have not yet been observed.

      I dont think you are right, because there is at least one model which count with two hIggses, oscillating against each other as the transport mechanism of information the origin of Casimir effect and filling lattice space at the planck scale. Cheers, Leo Vuyk.

      2013/7/6 Of Particular Significance

      > ** > ohwilleke commented: “”No one will be surprised if there is a second > type of Higgs particle in nature. A very wide variety of speculative ideas > that go beyond the �Standard Model� (the equations that describe the known > particles and forces) predict the existence of more t” >

  7. I’m told that the idea that, when the title is a question, the answer is no, is known as Hinchliffe’s Theorem. 🙂

    1. Heather–Actually, Boris Kayser wrote a one-page paper once (with byline: Boris Peon) entitled “Is Hinchliffe’s Theorem True?” The paper said “Hinchliffe has proposed that whenever the title of a paper is a question, the answer to the question is “no”. This paper shows that HInchliffe’s Theorem is true, but only if it is false.”

  8. By deleting my remarks about the masses of three neutrinos, I am assuming you feel they have no bearing on the masses of higgs ? It was because the three are in the ratios 1,2,3 that one can predict the mass of higgs2 to be 142.016. Please feel free to delete again.

    1. Aside from the fact that it is rude and inappropriate to mis-use another person’s website as personal advertising space, no professional could predict the mass of the Higgs particle to 6 decimal places, because there are processes that shift the mass of the Higgs particle whose sizes are not currently known to that accuracy. Your very prediction reveals your ignorance of basic issues in quantum mechanics, because you have clearly not accounted for quantum corrections to the mass of the Higgs before you made it. Since you don’t even know the basics, why should we trust you?

      1. Having some soft of “more” button on long posts would be appreciated, especially for “new age” or “self taught but misunderstoon physics genius” types. I find it anoying to scroll past lots of waffle. Of course if you can edit the most egregious posts before we see them, that is even better.
        Of course Matt, all of your sentences are worth reading!

  9. What is change on higgs spin , decay , if there is new one higgs (second Higgs) and is this change the last observation discovery of higgs boson?

    1. Higgs particles’ spin is zero, period.

      Higgs decays can indeed be modified if there are multiple Higgs particles, as can be production of the Higgs particles. Data is king, so clearly any changes due to the existence of other Higgs particles cannot be so big as to be in conflict with what the data currently says. But the data does not say the Higgs that was found last year is a Standard Model Higgs particle. It only says that within about 20-30%, some of its decays and production rates resemble that of a Standard Model Higgs. So data certainly allows for substantial — but not enormous — differences between the particle we have observed and what we predict for a Standard Model Higgs particle.

    1. They have and the new particle is currently being detained in Guantanamo bay. This is why they had to spy on their European allies.

  10. “If the bottom quark or the muon serve a “purpose”, we have no idea what that purpose is. And we’ve known about these particles for decades.” Could the purpose of the 3 generations of particles be to allow D-brane charges to exist?
    http://arxiv.org/pdf/0909.0689v3.pdf “Topics on the geometry of D-brane charges and Ramond-Ramond fields” by F. F. Ruffino, 2009
    In order to describe 1 generation of particles, one might need 3 dimensions of linear momentum, 3 dimensions of angular momentum, 1 dimension of quantum spin, and one dimension of time. Could there be 3 generations of particles in order to enable the Leech lattice to play an important role in string theory?
    http://en.wikipedia.org/wiki/Leech_lattice

  11. There’s nothing quite so rum as leftover news warmed up to be something to chew on today.

    What is the probability that the first bump isn’t a statistical fluke and that the absence of a second bump is? (That is there should be two bumps but sheer random chance has hidden the second.)

    1. The first bump would not be enormously significant even now, probably in the 5-6 sigma range combining ATLAS and CMS, but you should remember that it is supplemented by bumps in the same place in the search for Higgs –> 2 lepton/antilepton pairs, which push things way up toward 10 sigma. (There is no bump at 136 GeV in those searches, in case you are wondering… but you wouldn’t expect one, for reasons that are too long to explain here.) To this we can add a signal in the decay to a lepton, antilepton, neutrino and anti-neutrino — though this doesn’t give a precise mass measurement.

  12. Donofexistence:
    Yes. I agree. I was kind of careless in wording. But if they do find more than one Higgs particle with different masses and different other properties, a difficult question will remain as to how they are related to a common field which gives rise to masses. Perhaps people have already speculated on this.

    1. Yes, Kashyap, that will be a difficult question. I look forward to hearing some of the thoughts on this.

    2. This is a long story — complicated, though not that difficult technically. But the short answer is: people have been speculating for decades; there are many, many ways to introduce more than one Higgs field; and in general, in these theories there isn’t a single common field that gives rise to masses. In general, masses of the known elementary particles may come from one field, or they may come from multiple fields. There are many, many possibilities. The leptons may get their masses from one field, the quarks from another. The up-type quarks may get their masses from one field, the down-type quarks from another. Some fields may not contribute to any masses, except perhaps the mass of the Higgs particles. And the Higgs particles are not so simply related to the Higgs fields, because they tend to be ripples in not one but some mixture of two or more Higgs fields. Still, the more Higgs fields, the more Higgs particles. See for example

      http://profmattstrassler.com/articles-and-posts/some-speculative-theoretical-ideas-for-the-lhc/supersymmetry/supersymmetry-what-is-it/

      http://profmattstrassler.com/articles-and-posts/the-higgs-particle/implications-of-higgs-searches-as-of-92011/

      The second one is out of date but contains examples that are still relevant now.

      I could go through an example or two in a future article, but it’s way too long to go through it here.

      But do not fear: this has been studied very, very extensively. In fact, the book “The Higgs Hunter’s Guide”, first published in 1990, has several long chapters devoted to these subjects.

  13. kashyap Vasavada wrote:
    But what other purpose apart from symmetry breaking and giving masses would these other Higgs serve? Perhaps it is very technical.

    In the above comment the implication is that other Higg’s particles give masses. However, from reading this blog and my understanding of how particles obtain mass, it is the Higgs Field that gives the mass to particles not the Higgs particle. The Higg’s particle is solid emperical evidence that the Higgs field exists. Seems to me with all the recent talk about the impreicision by journalists who talk about theoretical physics, we should strive to be more precise in our language about what is going on when we discuss particles, mass, and fields.

  14. In the supersymmetric or more complex Higgs models, are there any relationships between masses or all the masses are free parameters like the first Higgs mass? I would think that if they serve similar purpose, the masses should be close. Right?

    1. The answer depends on the theory, and even then, there’s no simple answer. And the fields involved don’t all serve the same purposes. So I’m afraid I can’t give you a one-line response. But here’s just one of a huge variety of examples: it could easily happen that there is one Higgs (the one we’ve found) with a mass of 125 GeV/c^2, four with a masses in the range of 450-500 GeV/c^2, and two more with masses of 3-6 GeV/c^2. You notice that they can be heavier or lighter than 125 GeV, and can come in groups, but there might be none with mass near to the one we’ve already found.

      1. Thanks. But what other purpose apart from symmetry breaking and giving masses would these other Higgs serve? Perhaps it is very technical.

        1. Does everything serve a purpose?

          If the bottom quark or the muon serve a “purpose”, we have no idea what that purpose is. And we’ve known about those particles for decades.

          Any extra Higgs particles might just be there because, well, they’re there. Or they might serve a purpose we haven’t thought of yet. Theoretical physicists have to consider things that have no obvious purpose, as well as those that do. You’d feel stupid if you failed to discover something for a few centuries because you’d never thought about it.

    2. The best model is the simplest model. Higgs2 is 142.016 Gev; Higgs1 is the only other one. [Abriged by host: this site is not intended to be advertising space for individual people’s ideas.]

  15. probabily the others bosons of higgs are in the extradimensions,or others spacetime continuos ( with other values to the light speeds as constant and limit),so as others antineutrinos that don’t exist in the spacetime continuos;because some neutrinos couldn’t to be revered in PT.

  16. The Higgs2 is at 142.016 Gev. = Higgs1*Z/W. The data shows a small bump there already.

  17. “John Ellis in his Higgs Symposium talk specifically discussed the impact of these measurements on a subclass of variants of Minimal Supersymmetry ( the “Constrained Minimal Supersymmetric Standard Model”). … Overall, my read is that only a small fraction of the variants of this subclass of models is still consistent with data.”
    http://profmattstrassler.com/articles-and-posts/the-higgs-particle/taking-stock-of-the-higgs-jan-2013/5-supersymmetry/
    Are there infinitely many non-minimal variants of supersymmetry that are physically plausible?

  18. The same statistical fluke in two different detectors: more than doubly unusual it would be unusual squared 🙂

        1. Oh, he won’t be surprised. There have been papers about the possibility of more than one type of Higgs particle for decades. Supersymmetric theories, for instance, have a minimum of five.

            1. The final outcome is unpredictable.

              But right now, the current data contains no evidence of anything other than one Higgs. And so we have no reason to think, even if there is a second Higgs, that it has a mass of 136 GeV.

  19. 3. Don’t forget, however, that a bump in the two-photon data from ATLAS and CMS wouldn’t necessarily imply a new Higgs particle anyway. There are other types of particles that could give bumps in this data!

    How, then, would you tell the difference between a new Higgs particle and other types of particles? Would the “other types of particles” be associated with the Higg’s field or some other field?

    1. Roughly speaking, for every field there is a corresponding particle… so a non-Higgs particle would certainly be from a non-Higgs field.

      Characterizing a new particle takes a lot of work. We would measure how often the new particle is produced, determine what other particles (if any) are simultaneously produced, and try to figure out into what other particles the new particle can decay — other than two photons. We would measure its spin and other more subtle properties. We would measure how much motion-energy it tends to have when it is produced. All of these things would contribute to our understanding.

      Indeed, we’ve been doing all these things for the particle we discovered last year. As a result of those studies, what we first called a “Higgs-like particle” in July 2012 (because it seemed a bit like a Higgs particle, but we certainly weren’t sure yet) is now, as of March 2013, called a “Standard-Model-like Higgs particle” (because we’re quite sure now it is a Higgs of some type, but we’re not yet sure it is of the simplest possible type.)

Leave a Reply

Search

Buy The Book

A decay of a Higgs boson, as reconstructed by the CMS experiment at the LHC

Related

Recently, the first completed search for what is nowadays known as SUEP — a Soft-Unclustered-Energy Pattern, in which large numbers of low-energy particles explode outward

POSTED BY Matt Strassler

POSTED BY Matt Strassler

ON 03/15/2024

About a month ago, there was a lot of noise, discussion and controversy concerning CERN‘s proposal to build a giant new tunnel and put powerful

POSTED BY Matt Strassler

POSTED BY Matt Strassler

ON 03/08/2024