> While exciting, there are countless examples in history of such discoveries disappearing as more data is collected.
As someone who works on ATLAS and has been watching this result develop over the last several weeks (without knowing what was going on at CMS), my excitement has been very tempered by this fact. Especially given that the global p-value is quite modest (see also: Look Elsewhere Effect [1]).
For sure it's too early to know anything. But I have to say, when I heard the CMS announcement yesterday that they were also seeing it, in almost exactly the same place, it's much more exciting!
On the optimistic side: has anyone done an estimate of the global significance taking into account the fact that CMS and ATLAS both saw the excess at the same place (~740 GeV)? (Clearly, if the situation was a simple as each bin being "excess" or "no excess", then you would only correct for the look-elsewhere effect for one of the experiments. When you turned your attention to the second experiment, that bin would already be special. This is muddied by the fact that things are not so clean, and the small blip of CMS was just the largest of several blips, etc.)
On the pessimistic side: has anyone done a "fully global" significance test, taking into account all the other bump hunts of similar a-priori interest?
> has anyone done an estimate of the global significance taking into account the fact that CMS and ATLAS both saw the excess at the same place (~740 GeV)?
Not yet, and it's actually a bit tricky to think about how this should be interpreted. Since neither experiment alone provides a super strong motivation to look at 750 GeV (since their global significance is low), it's hard to argue that one experiment can remove the LEE from the other by fixing the search mass to 750 GeV. But IMO it doesn't even make sense to do this thing post-hoc; otherwise one could always avoid the LEE by "fixing your attention" on a single hypothesis mass where you had previously observed a local excess.
Now, suppose you had a (probably contrived) theory which specifically predicts a particle at this mass in a way that is totally non-negotiable, so that the only meaningful question to ask about the theory is "is there a particle X at 750 GeV?". In this case you can do away with the LEE and use the stronger local p0 at that mass. Theorists are already flooding the arXiv with such articles (and even have been since a day before the announcement... oops!).
But in the general case, really the best way to understand the combined significance is to combine the datasets from CMS and ATLAS in a fully correlated way that handles systematics etc, and then make the diphoton mass plot again with the bigger dataset and evaluate the global significance of the resulting bump, still accounting for the LEE. But this is a really big effort, since the two detectors are substantially similar/different in many ways and it takes great care to handle correlations etc between their resulting data. Something like this was done for the Higgs discovery [1], in a paper which you might recall made news [2] for having more authors than any paper in history!
A coincidence is the most probable explanation for the
surprising bumps in data from the collider, physicists
from the experiments cautioned, saying that a lot more
data was needed and would in fact soon be available.
So the data analysis is extremely complicated, in part because the number of tries depends on the correlation of datapoints. So if you look at 2 sigma signal in one hundred datapoints you expect to find 5 and even worse, if you try one hundred different measures then you expect 5 two-sigma events, if the measures are independent, which they are not in practice. So to calculate the number of tries is a highly non trivial business, which depends on exact knowledge what each desperate grad student did on his weekend.
So the nice thing about data from both CMS and ATLAS is, that we can fix this by asking, what is the probability that CMS sees something at the same energy and in the same channel as the most interesting ATLAS result. And as it turns out, it is a 2-sigma event. So there is just a one in twenty chance that this occurs by random chance. However one in twenty is not exactly a discovery, just the probability that the Niners win on Sunday.
Edit: Two things, first that analysis is of course very conservative and second even if it is a real effect, it may be just a presently badly understood background effect.
Dumb question from non-physicist - if the 'Standard Model' is a good theory, why don't we have a good idea of where to find particles [ ie. predict their mass and other properties ]?
Is the problem that its just hard to compute these values ?
Or does the theory not predict particle masses ?
Or do we need a fully unified theory of all fundamental forces, before we can predict particle properties from first principles ?
Well, we have found all the particles predicted by the Standard Model, last one missing was the Higgs boson.
However, no-one really believes that the Standard Model is the ultimate theory. There must be other stuff out there that we have found yet. What people are looking for now are deviations from the standard model.
What do you mean? It predicted the Top and Charm quarks, the W and Z bosons (and their masses), and many other phenomena.
We know the Standard Model isn't a complete theory or a "grand unified theory" though because there are too many empirically determined constants in it, and because it doesn't adequately explain all known particle physics.
At the moment the limiting factor is energy -- The energies required to create these collisions (and hence, the possibility of these new particles) is enormous. In a way that's the only way the LHC exisists; to create higher energy collisions than before.
As for computing values: It's nice to have a theory, but it's nicer still to have it confirmed/falsified by expermiment.
Yes. The graviton in question would be a massive partner to the ordinary, massless one. You get those in Kaluza-Klein style theories with extra dimensions [1], so they are often referred to as "KK gravitons".
First I read "heavy boson" and thought that sounded kinda metal, so it would be cool if they named it that, but then I misread "whale boy", which is even better, so I'm calling it now. Next heavy higgs boson will be called Whale Boy.
> You see that the diversity of the explanations is huge – it's a landscape of possibilities. Whether you like it or not, whenever certain things are uncertain, the possible answers are diverse. String theory's landscape reflects the same principle in another, more consistent and more constrained, formalism. And the purpose of this new particle is simply mysterious at this moment.
> It seems very likely that the new particle is a j=0 scalar particle. And it seems guaranteed that it can't be the only new particle that has to be added relatively to the Standard Model. But who are the friends of SS and what is their relationship and purpose? Nobody knows. If this extra knowledge about the particle emerges, we may start to call it a heavy axion, a technicolor or composite scalar, a pseudo-Nambu-Goldstone boson, or – which I find more likely – a new Higgs boson analogous to the (three-year) old one.
"Parked along the underground racetrack are a pair of mammoth six-story conglomerations of computers, crystals, wires and magnets: Atlas and C.M.S., each operated by 3,000 physicists who aim to catch and classify everything that comes out of those microscopic samples of primordial fire."
Is that a typo? Are there really 6,000 physicists actively doing work on the large hadron collider?
There are far more, in fact. Each of these collaborations have around 3000 members, but there are also other experiments (LHCb, ALICE are the two big ones but there are some smaller ones too) and then there is also the team of accelerator physicists that operate the LHC.
Those are jobs for the CERN facility alone (which also probably includes services like HR, translation, janitorial, etc). But thousands more work essentially full time as "CERN users". These scientists and engineers are typically paid by member organizations around the world, i.e. universities (think professors, postdocs, grad students) and national labs (LBNL/Brookhaven/Fermilab etc in the US, in2p3 in France, DESY in Germany).
The ATLAS and CMS author lists are about 3000 each (generally all physicists), but all told there are over 10,000 scientists, technicians, and engineers who basically work full time running the LHC and its associated detectors.
That's not to mention all the people working on any of CERN's many other projects that have nothing to do with the LHC.
I respectfully disagree.
1) The Manhattan project is not comparable in scope: Both the UK and France developed nuclear weapons on their own.
2) This is not a European project per se. Most of the developed nations in the world contribute, and as such this is very much a project for and by humanity itself.
So to explain the world we see, we combine quantum theory, theory of relativity, dark matter, gravitation, electromagnetism, two types of nuclear force and a zoo of now 62 elementary particles?
I often wish for some kind of "Physical Philosophy". That tries to make sense of the world independent from those theories. For example: what if the theories just keep growing indefinitely? If we keep "finding" more and more complex theories that explain some aspects of what we see better then the old ones and keep "finding" new particles? What would that tell us about reality? What if we found a very simple formula that explains everything with just one type of particle and one dimension? Would it matter?
The world is complex, but we haven't gotten to the stage yet where you should feel lost, even as a layman. It is absolutely possible to know enough classical physics, QM, and theory of relativity to have a decent understanding of our current models. You absolutely don't have to be able to recite a table of elementary particles in order to understand this.
> I often wish for some kind of "Physical Philosophy". That tries to make sense of the world independent from those theories.
If your sense-making framework can (or must?) be independent from the facts, there is an enormous multitude from all cultures available for you to choose from. But if facts are important to you, and you don't know where to start, watch some popular science documentaries. If you prefer a touch of philosophy, I can suggest the original Cosmos series by Carl Sagan. But on the whole, sense and meaning is not something that can be prescribed by scientific knowledge - it can, however, help inform your own decisions about sense and meaning.
> For example: what if the theories just keep growing indefinitely?
This does not describe the current state of physics research. In fact, we are almost struggling to discover new physics, and so far most particle physics experiments line up so well with theory it's almost frustrating. What we can see right now is not the bottomless fractal recursion into nothingness which you describe.
Look, its just that it's all so complex. These must be emergent phenomena of something deeper - otherwise you have a universe which pops into existence fully formed with these particular properties, this particular family of particles and forces. These experiments are a bit like 3d scanning a tree in millimeter detail and then declaring that you have a perfect knowledge of trees. Don't get me wrong this is very important stuff, we need this knowledge to find the underlying systems. But it creates a degree of philosophical unease in the sensitive individual. For me it seems necessary to say that if this universe is all that exists, it must necessarily exist from first principles. My hunch is that what we are seeing is the population dynamics of all possible systems of interacting elements - or atleast a tiny little piece of it.
> These must be emergent phenomena of something deeper
Deeper might be the wrong word in this context, but I don't think many physicists would disagree that there is probably a small set of rules and variables that makes up the substrate of reality. For all we know, the universe could be a really basic cellular automaton.
> These experiments are a bit like 3d scanning a tree in millimeter detail and then declaring that you have a perfect knowledge of trees.
That does not describe our current scientific process at all, which is to find underlying abstractions and models instead of just cataloguing concrete findings.
I also think there is a problem with the understanding of higher-order (emergent, as you say) phenomena: they often do not follow from the basic underlying rules in an immediately obvious manner. What they do is they become systems in their own right, with their own mechanics worth studying.
To use an analogy, just because you know everything about sugar and flour doesn't mean you know everything that's going on in a bakery.
> But it creates a degree of philosophical unease in the sensitive individual.
It's not the universe's job to make us feel comfortable, nor is it science's job to discard findings on the grounds that we personally don't like their implications. The universe does not care about our comfort, and philosophical considerations are not part of the fabric of nature outside of your personal mental frameworks.
There are many things about nature or the universe at large that make me uncomfortable, but that doesn't mean these facts should somehow be discarded.
> My hunch is that what we are seeing is the population dynamics of all possible systems of interacting elements - or atleast a tiny little piece of it.
I read this several times, and I have no idea what that means.
The universe is what it is, and may not be to anyone's taste. No one is declaring perfect knowledge. If anything, experimental physicists are much more humble than theoreticians (let alone philosophers) and are interested in teasing out the truth without regard for its conceptual elegance.
That's pretty much exactly what theorists are trying to do. For example, we used to think that electricity and magnetism were different forces and subject to different equations/rules. Then we united them into a theory of electromagnetism and realized they're governed by the same equations. Next we discovered the weak nuclear force which seemed totally different, only to later realize that the weak and electromagnetic forces are actually both part of the same (electroweak) force, and both are governed by the same fundamental equations.
It is suspected (and/or hoped) by physicists that this trend continues; that the other nuclear force can be united with the electroweak forces under a single theory; there are already hints that this is the case (see: Grand Unified Theories).
So far not much progress on getting gravity to fit in with the picture though :/
Having things be complex and messy is hardly ideal, but I think Neil deGrasse Tyson's sentiment is important to keep in mind: The universe is under no obligation to make sense to you.
But as cshimmin mentioned, we already unified electricity and magnetism. I am reasonably confident we'll find something equally nice and elegant for the rest. Maybe not tomorrow, or in a hundred years, but it'll happen. If only because we're too stubborn a species to let it be.
>So to explain the world we see, we combine quantum theory, theory of relativity, dark matter, gravitation, electromagnetism, two types of nuclear force and a zoo of now 62 elementary particles?
Well, the standard model includes all the forces and particles, and neither the particles or the forces are arbitrary, but arise from various symmetries in the laws of physics.
The symmetries in the standard model have been so good at predicting and explaining particles and forces that many physicists are convinced that there are a whole class of undiscovered particles to go along with another type of symmetry.
They didn't build the LHC just to see what's out there. They have pretty strong reasons to believe that there are these other particles. The fact that they can predict these particles (like the higgs) should tell you that there is a underlying theory that explains all of this.
The LHC is mostly just helping the physicists discover the details about the particles that aren't predicted by the theory (such as the mass).
Simple workaround. Copy the article title and paste it into a Google search. When Google is the referrer, they drop the paywall [1]. For NYTimes and WSJ articles, this works almost every time.
Uh... the NYT is a news outlet. They reported news on the fact that we had a conference, and summarized the contents of the conference (we see possible hints of a new particle). It's not like we published a paper in a scientific journal claiming anything of the sort.
At present there is also a write-up on the homepage of the NYT titled "NATO Nations No Longer Question Need for Alliance," which presumably doesn't reject any null hypothesis either. I guess we should snark about that as well.
They're at about a 90% confidence level. They prefer to have better than that by several orders of magnitude, but it's probably worth talking about it a bit publicly at this point.
There are lots of theoreticians writing papers right now trying to figure out what it could be. Most likely seems to be a second higgs (lots of theories predicted several flavors of higgs).
I believe "There is no such thing as toast." is only a theorem.
Theorems may be rejected all day long.
Edit: A null hypothesis in this schema would be something like (a bit ridiculous, but not my example): If toast exists and I eat it for breakfast, I should be able to detect that I ate it for breakfast at a higher rate than the false positive rate of my ability to determine whether what I eat for breakfast is or is not toast.
Keep downmodding. Or, you know, refute what I have said.
the term "null hypothesis" usually refers to a general statement or default position that there is no relationship between two measured phenomena, or no difference among groups
Rejecting or disproving the null hypothesis — and thus concluding that there are grounds for believing that there is a relationship between two phenomena — is a central task in the modern practice of science, and gives a precise criterion for rejecting a hypothesis.
I didn't vote on your comment but I can't make enough sense out of it to write a refutation. There is a sort of "evolution is only a theory" flavor to the whole thing.
https://atlas.web.cern.ch/Atlas/GROUPS/PHYSICS/CONFNOTES/ATL...
http://cms-results.web.cern.ch/cms-results/public-results/pr...
This video shows how one of the plots evolves over time as you discover a particle (in this case, the Higgs, discovered in 2012):
https://www.youtube.com/watch?v=zLZZOrpQFo4#t=3m
A nice blog post:
http://resonaances.blogspot.co.uk/2015/12/a-new-boson-at-750...
While exciting, there are countless examples in history of such discoveries disappearing as more data is collected.