▲ 189 points
back
113 comments
The fundamental theories we have is Quantum Electrodynamics and Quantum Chromodynamics.
This new experiment tells you that "modern nuclear forces, including those derived within chiral effective field theory" break down and cannot be used to describe what they observed. Here is the arxiv: https://arxiv.org/abs/2112.10582
It just tells you that their effective theory is no longer effective in these circumstances. Unless you actually observe something contradicts QED+QCD calculations, nothing fundamental is wrong.
I can't believe they pick such a clickbaity title for a serious publication and let quanta magazine publish an even more clickbaity article about it. Well, I guess they need more funding.
[EDIT] PS. The science is sound and suggests that nuclear physicists must refine their theories to match observations. This also encourages those working on QED+QCD, as increased computational power may enable precise form factor calculations for comparison with experiments.
The title says that the experiments diagree with "the leading theory of the nucleus", which is chiral effective field theory. I don't see anything wrong or "clickbaity" about this.
I would characterize it as the article doesn't have anything overtly wrong with it, but it did make it easy to come away thinking this is revealing some sort of massive fundamental issue with everything, rather than what it is, which is the continuing process of tuning a highly successful theory down in the decimal points.
Whether or not you consider that to be part of their job is in the eyes of the reader.
Small discrepancies in the observed orbit of Mercury ended up changing our understanding of space and time in most profound ways.
Sorry, but unless you're a physicist, why would you expect to be able to know the significance of the paper? If you walked away misunderstanding the implications, then I gotta ask - assuming you're in software - are you careful to proofread your papers to ensure a geologist doesn't misunderstand them?
"Theory of the nucleus" does not necessarily imply just QED+QCD in this context. It can mean something derived from QED+QCD - the theory of which elements of the calculation can be thrown out from the QED+QCD calculations. That would be hardly the first theory which is a derived theory from a more fundamental theory. So the headline is fine? For once Quanta was less clickbaity than the preprint.
Thanks for the arxiv link.
I think you're reacting to things the paper doesn't say - they never claim any new physics. EFT is a simpler model than what you'd derive from QCD and it makes sense it'll be wrong in some limits. The paper finds such a limit. That's all? The paper title and abstract seem to be pretty accurate? Or are you objecting to the characterization of EFT as "theory"?
If so, then how about planet formation theory, or solid state physics? Are those also not theories, because they're ultimately just limits of the standard model?
A title of "theory mostly sound, except some extreme edge cases," would cover about 95% of science research.
You think that 95% of science research is news worthy?
> I can't believe they pick such a clickbaity title for a serious publication and let quanta magazine publish an even more clickbaity article about it. Well, I guess they need more funding.
My take from reading the article (not a physicist) is the physics of the nucleus (i.e. the protons and neutrons and strong nuclear force) are treated as emergent phenomena from the quantum theories about fundamental particles (e.g. quarks, leptons & gluons).
Per the article, such leading theory of the nucleus is "chiral effective theory," which seems to be quite inaccurate at making predictions for the experiment in question.
I'm not sure how much more accurate the headline could be, unless chiral effective theory is not in fact the ex ante leading theory of the nucleus.
> "Well, I guess they need more funding."
are they trying to clickbait jim simons
why do we clutch pearls whenever a publication uses a "click-baity" title? at least it manages to a: get people to actually read the publication, and b. publisher/site makes money. nobody's running a charity of information in a capitalist society.
Having an standard of truth is important as a reputation for an organization(whether for commercial or non-commercial reasons). Besides, click-baitiness is not sustainable as in the equilibrium, readers will change expectations to assume artices are click-baity and information on the real breakthroughs will have fight all the fake breakthroughs.
BTW, This is not about Quanta which I find is a really good resource. Although, there might indeed be some click-baitiness not with the motive of profit(you could describe it as a charity run by Jim Simons), but in the sense of someone trying to get students interested with a provocative title for a science talk.
agreed. It's a fine line that publishers have to walk between capturing attention and maintaining credibility. As you've said, the long-term sustainability of clickbait is questionable. It's a short-term strategy that may inflate traffic numbers, but can lead to a loss of trust over time.
The challenge is to strike the right balance between making content appealing and keeping it accurate and valuable, especially crucial with scientific publications, where the accuracy and trustworthiness of information is paramount.
> Instead, at these scales, a new effective force emerges — the strong nuclear force, transmitted between nucleons by the exchange of pions.
I'm admittedly ignorant, but I'm not convinced that a lot of theoretical physics is basically curve fitting where you've got a model with enough flexibility that you can make it fit the data. String theory always seemed like an egregious version of this, I think it's less popular now.
I think that's why elegance goes a long way in theories: a simple and concise description is harder to overfit.
The standard model has a number of free parameters [1] but the results of calculations are pretty sensitive to them, and this is what predicted (not just explained post-hoc) the Higgs. It's not incorrect to say there's curve fitting involved, but it severely understates the magnitude of the achievement that is the standard model, and its success in explaining everything that underpins everyday life[2], as well as most things beyond that regime[3]
[1] https://en.m.wikipedia.org/wiki/Standard_Model#Construction_...
[2] https://www.preposterousuniverse.com/blog/2010/09/23/the-law...
[3] https://physics.stackexchange.com/questions/128374/evidence-....
The standard model doesn’t explain GPS satellites.
But relativity explains GPS. And the two theories together explain practically everything to insane levels of precision.
The only time they go wrong is when they both operate at once, in a narrow range around black holes or the very earliest universe.
If it explains everything then why are experiments still ongoing? What is left to explain?
Experimental physics works a bit differently than most people think. While string theory was/is popular, very few hold it as certainty. From the very beginning the discussions around it were that it was a non-testable theory. While out-group conversations around physics are "physicists believe" in-group conversations are more "the leading theory is". Uncertainty is deeply ingrained in the study and the focus is more around making our knowledge less wrong rather than proving something correct/making our knowledge correct. Subtle, but notably different. It is the reason you always see quotes with qualifiers like "should" or "we think" and rarely certainty (easy to miss if you aren't looking).
These distinctions matter when you create experiments and interpret results. They generally don't matter to the layman, since an authoritative answer is good enough. But they do matter if you need to do any form of evaluation, as essentially what I'm talking about is the importance of including error and uncertainty. Which btw, particle physics often has a uniquely tight bound: 5 sigma. You'll even notice CERN's blog post about 5 sigma has lots of qualifiers, does not suggest it claims certainty, how it isn't alone enough, and even references that there are good arguments for even higher bars. It's a different language than people are used to.
I don't get why the arguments are confined to string theory. The standard model plus general relativity are so good that any theory that resolves the conflict is necessarily only testable when you've got a stellar mass black hole in your lab.
That could be a call to stop bothering with exploring the domain entirely as infeasible. But I don't understand why string theory gets singled out for finger waggling when any other theory must run into the same problem.
What passes the bar for "elegant enough"? Why would laws of nature care about such human principles? And, most of all, why treat elegance as a guiding principle when all it has done is produce models incapable of being accurate outside a narrow range?
Historically, haven't the more "elegant" models been shown to be correct? The various explanations for planetary motion come to mind as an example.
You may enjoy The Structure of Scientific Revolutions by Thomas Kuhn. Fantastic book.
Reminds me of Freeman Dyson's anecdote of Fermi:
"I remember my friend Johnny von Neumann used to say, with four parameters I can fit an elephant, and with five I can make him wiggle his trunk." [1]
The theory they’re talking about is not just curve fitting. Instead only pieces which match the physics of QCD / the Standard Model Are included. It is simple to write down terms that you might include in a “just make up enough terms until it fits” methods that are excluded based on the requirements of symmetry matching that the whole program of EFT is organized around.
Particle physics is probably still stuck in the butterfly collecting phase, in which case “curve fitting” is not a criticism.
I think the "butterfly collecting phase" in particle physics was really the 40s and 50s, when a whole bunch of surprising new particles were discovered.
Gell-Mann and other proposed a unifying principle in the 60s, and as far as I know that's been broadly successful and ultimately led to the standard model. https://en.wikipedia.org/wiki/Eightfold_way_(physics)
We are not now discovering any new unexpected particles (or any theoretically expected ones, since the Higgs), which is a bit unfortunate in terms of giving theorists something to work with.
It's entirely that.
String theory looks like a good math framework to describe possible universes. Physics today is hindered by the obsessive idea that what we see around - the observable universe - is all there is, rather than just one of many possible worlds. This is the good old geocentric model on steroids. And this is why physicists are puzzled by the lucky combination of constants that define our world.
This is great news, if it holds up. There hasn't been a lot of remarkable advancement in physics since the Standard Model.
"Another topic that comes up is simplicity. According to Feynman, nature is usually much simpler than our thoughts. Therefore, when trying to explain phenomena, we tend to overcomplicate things. Often, in the end, reality can be explained by much simpler terms. We just need to look at it from another point of view."
- https://cassandradispatch.org/richard-feynman-on-looking-at-...
The problem with the nucleus is that the internal dynamics there are impossible to describe with any reasonable precision.
If for the electron shell we can calculate pretty much anything we want from the first principles (energy spectra, the half-lives of unstable and metastable states, etc.), we can't do anything similar for the nucleus.
For example, we can't compute half-lives of unstable nuclear isotopes. The best models are on the level of "imagine that a nucleus is a drop of water" or "assume that a nucleus is a potential well that contains an alpha particle".
And no, this is not a fundamental theory issue. We can describe the behavior of individual nuclear particles just fine at the energies that exist within the nucleus. It's their interaction that is completely baffling.
> If for the electron shell we can calculate pretty much anything we want from the first principles (energy spectra, the half-lives of unstable and metastable states, etc.), we can't do anything similar for the nucleus.
As I understand it, we can only really do that for a single-electron atoms. Multi-electron interactions get the same problem as inside the nucleus.
It doesn’t sound like this is necessarily a challenge to the Standard Model, only to one of the ways of approximately deriving predictions about the strong force.
For real, I wonder if Physicists are not putting themselves in a corner by using equations that are in principle "easy to work with" (because they're pulled from similar phenomenon) but a pain to actually use
Feynman diagrams are a visual way of representing a boatload of very complex equations and the worse part is that they work! But maybe it's a failure of math more than physics
How do you excite a helium nucleus?
It sounds like the setup to a joke, the punchline being that you tell it there's cake in the dining room, but I am honestly curious?
Laser bombardment? Running it through a particle accelerator? Pointing out the girl helium over at the bar looked at it?
Ok, sure, they shot electrons at it, but why would that have any real, measurable effect on the nucleus?
What are the conditions that qualify as "excited" for a helium nucleus? Is it just the "balooning", or is there some real meaning to the term?
I have so many questions.
"It’s possible that simply including more terms in the approximation of the nuclear force might be the answer. On the other hand, it’s also possible that these ballooning helium nuclei have exposed a fatal flaw in our understanding of the nuclear force."
I bet on more terms.
I think that’s a fair bet; that’s pretty much what physics have been doing to every problem for the last 100 years unfortunately..
I am far from an expert. But from the article, it seems to me to be more likely that it casts doubt on the approximations they're using to do the calculations.
Well yes, but our understanding of what approximations should work is not really separable from the theory itself. E.g. from the article,
> Van Kolck contends that some of the parts deemed less important and routinely ignored are in fact very important.
So it's still good science IMO.
That's my understanding: the math required to get a concrete prediction of a model is complicated enough that “either the theory, or the experiment, is wrong” excludes the most likely possibility: that too much error is being introduced by the approximations that they're (necessarily) using to get a tractable computation out of the model.
Something something, all models are wrong, some models are useful.
if we have to come up with a completely new model, we should at the same time reduce confusion going forward by renaming the new model to "the nucular model of the nuculus".
Nuculus. It's pronounced nuculus.
I am not a physicist but if I were to take a stab at simplifying things I’d assume that a neutron is probably just a proton and an electron… isn’t that what a neutron decays to once it has been ejected from a nucleus anyways?
I mean, it's not a bad theory. The wikipedia article on neutrons discusses how this was an assumption somewhat like this about 100 years ago ("nuclear electrons") and the lines of reasoning that led to abandonment of the idea: https://en.wikipedia.org/wiki/Neutron#Discovery -- it has to do with the observable quantity called "spin"; the spin of a neutron is not the sum of the spin of an electron and a proton, so it must be something else.
“The complexity of the proton and of the neutron seems to be real, and not due to a lack of knowledge on the part of physicists. We have equations that we use for describing quarks, anti-quarks and gluons, and the strong nuclear forces that they exert on one another. [These equations are called “QCD”, short for “quantum chromodynamics”.] We can check the accuracy of those equations through many different measurements, including the rates for producing various types of particles at the LHC. And when we put the QCD equations into a big computer, and make the computer calculate the properties of protons and neutrons, and other similar particles (collectively called “hadrons”), the computer’s predictions for the properties of these particles closely resemble what we see in the real world. So we do have good reason to believe that the QCD equations are right, and that our knowledge of the proton and neutron is based on the right equations. Yet having the right equations isn’t enough by itself, because
• simple equations can have very complicated solutions, and
• sometimes it is impossible to describe complicated solutions in a simple way.
As far as we can tell, that is the situation with nucleons: they are complicated solutions to the relatively simple equations of QCD, and there seems to be no way to describe them in a few words or pictures.”
-- https://profmattstrassler.com/articles-and-posts/particle-ph...
Probably the most straightforward and stark demonstration of why a neutron can't be just a proton, electron and neutrino all stuck together is that it entirely fails to explain where all the rest of the particles come from when you have a jet: https://profmattstrassler.com/articles-and-posts/particle-ph...