Articles, with very little detail, have made their rounds about an X17 boson (16.7 MeV) being observed in tests of decaying beryllium-8 and perhaps once in a test with helium.

Most of the undiscovered particles that are searched for in CERN or other colliders or dark underground quiet dark pools filled with water were predicted prior to being searched for. (WIMPs, the Higgs boson, neutrinos, axions - all predicted first), after which, there was a long period of trying to find them, in some cases, still ongoing.

And, likewise, sometimes objects are observed before they're predicted (dark matter and dark energy comes to mind, probably more).

I know that a 5th force has been theorized, so in a sense, a 5th boson has long been proposed, but the articles I've read imply only that beryllium decay products were studied and X17 has been observed as a bump in the chart of output data, but no indication of what they were looking for when studying beryllium decay. I realize the discovery isn't official yet, but some articles suggest pretty good certainty that the observation is legitimate.

No mention of what they were looking for suggests to me that this was a surprise discovery, not a study where they expected to find something new, but I wanted to confirm, as the articles I've seen are all very short. If there was something specific they were looking for when studying beryllium decay, then identifying that would be appreciated.

A paragraph or two in layman's terms is also appreciated.

  • 7
    $\begingroup$ are we pretty certain that it is legit? "The same group had claimed to find various other new particles earlier in 2016, and abandoned these claims later, without an explanation of what caused the spurious signals. The group has also been accused of cherry-picking results that support new particles while discarding null results." source $\endgroup$
    – Michael
    Nov 28, 2019 at 19:01
  • 1
    $\begingroup$ @Michael But another team claimed to have found X17 too. So, no, but it's enough to talk about it. $\endgroup$
    – wizzwizz4
    Nov 29, 2019 at 7:26
  • 3
    $\begingroup$ @wizzwizz4 No, it’s a followup study from the same team. $\endgroup$
    – knzhou
    Nov 29, 2019 at 18:33
  • 1
    $\begingroup$ @userLTK That news article illustrates an unfortunately common misunderstanding of $p$-values. A $p$-value isn't a probability of being a fluke at all. It's the probability of getting a result at least that extreme assuming the hypothesis is false and that there are absolutely no unknown systematic effects. $\endgroup$
    – knzhou
    Dec 1, 2019 at 4:28
  • 1
    $\begingroup$ You can get tremendously misled by $p$-values. For example, suppose I flip a coin and get the sequence "HTHHHTTTHHTHHHTHTHHTTHHTTHHHTH". Let my hypothesis be that there is a worldwide government conspiracy with the sole goal of making my coin performing that 30-flip sequence. The null hypothesis is that this sequence was just a fluke. The $p$-value is $2^{-30} \approx 10^{-9}$. This does not prove that a government conspiracy exists. $\endgroup$
    – knzhou
    Dec 1, 2019 at 4:31

2 Answers 2


Is the X17 predicted, or discovered? It depends upon what the meaning of the word "is", is.

Prediction vs. Discovery

In particle physics, phrases like "the Higgs boson" or "dark matter" actually stand in for mechanisms or effects, which could be realized in many different concrete models. And even when you specify a model, it's still ambiguous because you can vary how "fundamental" your model is.

The Higgs boson itself is a good example of this. Before the LHC, we knew for sure that particles in the Standard Model had mass, that this was difficult to accommodate the standard way (by adding so-called "mass terms" to the Lagrangian), and that a very simple and elegant way to do it was the Higgs mechanism. By calling it a mechanism, we mean that it required only general ingredients, such as spontaneous symmetry breaking and new bosons, but that it wasn't specified how the symmetries were broken or what new bosons were added.

What we call "the" Higgs boson is just the simplest set of ingredients you could supply to get the Higgs mechanism to function. There were many alternatives, such as the two-Higgs-doublet model, which would give you five Higgs bosons instead.

And even once you fix "the" Higgs boson, you could go deeper. In the resulting model, the mass of the Higgs boson is put in by hand, and basic consistency requirements only fix it within a rather wide range. You could consider more complex models that predict more specific values of the mass, or even explain where the Higgs boson itself comes from, e.g. if it falls out of an elegant grand unified model or is itself made of other particles, like the proton is made of quarks. There were so many of these theories that just about any Higgs mass in the reasonable range could be accommodated by one or ten.

So I would say it's fairly clear that a Higgs boson was predicted before discovery... but it's completely up for debate whether a Higgs boson of mass $125$ GeV was predicted or discovered!

Fifth forces

Like "dark matter" or "the Higgs boson", "the fifth force" is another one of those vague phrases that really stands for thousands of distinct models -- basically any model that consists of the Standard Model and any new gauge bosons. Fifth forces are interesting because they are extremely simple extensions of the Standard Model (e.g. you could just take one of the existing forces and exactly copy it), and many more complex/fundamental models give you them automatically. Another appealing feature is that many types of fifth forces give rise to striking experimental signatures, which we can look for with great sensitivity.

So in that sense, a fifth force has been predicted for a long time. However, because of the avalanche of possible distinct fifth force models, experimentalists usually don't try to test specific models. Instead, they parametrize the effects of the fifth force in terms of a few quantities (the mass of the gauge boson, the gauge coupling, the coupling to electrons, the mixing with the photon, etc.). Then they run tests that try to capture as much of this parameter space as possible. Many such tests have been performed in the past and are ongoing now.

The experimentalists behind the X17 boson were looking for a specific type of fifth force, called a "dark photon". These are part of a relatively new mechanism to produce dark matter (DM), using an entire "dark sector" where the DM only affects normal matter through the dark photon. This mechanism has been gaining in popularity because it results in lighter DM, and the traditional heavy DM has been pretty conclusively tested by WIMP searches. However, "dark photons" are still not a specific model, but rather stand for a general idea that has a range of corresponding concrete models, and within each one the mass of the dark photon can lie in a wide range. So they were inspired by these theoretical ideas to look for MeV-scale bosons, but not directed to any specific mass; they just were trying to capture as big a slice of parameter space as possible.

They indeed found evidence for a new MeV-scale boson, but it behaved in a rather strange way. In a later theory paper, it was explained that the boson could not have been a dark photon, as a dark photon with that mass and coupling would have already been discovered in earlier experiments. Dumping the dark photon interpretation, the theorists showed that the observations could be accommodated, without conflict with earlier experiments, if the new boson had the unusual property of being "protophobic", i.e. not coupling to protons. The experimentalists then did a follow-up work confirming their results, leading to the recent media frenzy, and that's where we are now.

I hope it's apparent that your question does not have a yes or no answer! Fifth forces have been thought about for a long time, but they're a very general thing. The experimentalists were motivated to look by a recent, particular kind of fifth force (dark photons) -- but what they saw wasn't compatible with it. Science often progresses this way.


I realize the discovery isn't official yet but some articles suggest pretty good certainty that the observation is legit.

This needs to be said: I don't think any practicing particle physicist would give this observation more than a 5% chance of holding up. At any moment, there are about 50 distinct experimental anomalies, each of which could revolutionize physics if true. Historically, the vast majority don't pan out. That's because physics is subtle, experimental physics is even more subtle, and these measurements are often pushing the limits of what our equipment can do. (There is the further problem that the lab claiming the X17 has a history of reporting similar anomalies -- if I recall correctly, they already pulled this exact same thing twice in the past, with different boson masses, and never explained why those observations went away.)

It just so happens that you've heard more about this anomaly because it blew up in the media. This happens naturally: the more prominent a story gets, the more each journalist wants to write their own take on it, and so you automatically get a winner-take-all situation where X17 gets more attention than the next ten most recent anomalies combined.

I wouldn't recommend worrying about the X17. Science will go on, the experiments will be thoroughly scrutinized and replicated, and in ten years' time we'll know for sure if it's real or not. If it is, you'll hear about it everywhere. If you hear nothing, assume it disappeared.

  • 1
    $\begingroup$ What problem are these fifth forces models trying to solve? "basically any model that consists of the Standard Model and any new gauge bosons". Why add new guage bosons? $\endgroup$
    – thosphor
    Nov 28, 2019 at 14:46
  • 11
    $\begingroup$ @thosphor There are so many different reasons that it would perhaps be better to say that we have a long list of problems to solve, and new gauge bosons are one of the standard model building tools to solve them. For example, the dark photon example I gave has to do with how dark matter is produced. Other fifth forces come in trying to explain the tiny neutrino mass, or experimental anomalies (like this one). Often you don’t even want them specifically, but they fall out automatically from a larger theory made to solve a range of different problems. $\endgroup$
    – knzhou
    Nov 28, 2019 at 16:57
  • 3
    $\begingroup$ If [experimental_hypothesis] wasn't at the edge of what we can measure, then we'd have figured it out back when it was at the edge we can measure. $\endgroup$
    – Gloweye
    Nov 29, 2019 at 12:16
  • $\begingroup$ Kudos for defining “DM” before using it. Now how about “WIMP”? $\endgroup$
    – WGroleau
    Nov 29, 2019 at 16:52
  • 1
    $\begingroup$ @WGroleau "Weakly interacting massive particle," a catchall for dark matter candidates which are fundamentally quantum-mechanical. Contrasted with massive compact halo objects (MACHOs), dark matter candidates like rogue planets or isolated medium-mass black holes. Most MACHO candidates are inconsistent with things we have learned about dark matter since the 1990s; the literature is vast. $\endgroup$
    – rob
    Nov 29, 2019 at 19:47

What can I say? Was the Higgs boson discovered? The task of the Higgs boson was to confirmation of the higgs mechanism. Now let me ask this stupid question: What mechanism gave mass to the Higgs boson itself?



http://pinopa.narod.ru/ (In Russian and English)

QM standard model can not answer this question.

Was the x17 particle predicted? It seems to have been predicted. However, quantum mechanics did not do it. However, in this case, too, the particle found is not what it was meant to be in QM model.

The recipe for the formation of (more or less stable) particles can be found here:


Looking at the graph, there is still a whole particles to be discovered.However, there will be nothing interesting or exciting about them. You just have to stop looking at particles the way a cow looks at a barn - that is, without understanding.

Sorry for My English.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.