Friday, April 27, 2012

Columbia Higgs panel

Bill Zajc sent me a link to this 1-hour panel discussion about the God particle at the Columbia University. It was moderated by Amber Miller, a physicist and a science dean of Columbia's FAS. Brian Greene (string theory), Michael Tuts (LHC), Dennis Overbye (NYT), and Mariette DiChristina (SciAm) were the panelists.

Click the screenshot to play the video.

I have watched it and those people are of course good physicists or good journalists as well as good speakers and they have presented a lot of the standard comments related to the Higgs boson. But I still felt uncomfortable about some general features, the character of the content, and the tone of their comments. Some of the stuff was fine but much of it was frustrating to listen to because it wasn't morally true.

First, I think that if someone follows the events surrounding the LHC at least at the level of the reports that get to the media, he hadn't really learned anything new about the search for the Higgs boson. Most of the things they said could have been said 20 or 30 if not 40 years ago. Such a physics fan could attend an event in America's largest city with the most important people at the Columbia University who are doing these things but he wouldn't learn anything new.

Michael Tuts of ATLAS gave an OK introduction to the collider physics, detectors, and said a few words on inverse femtobarns etc. But when it came to the question whether they had actually seen the Higgs, he actually took the position "we are completely uncertain" that Matthew Strassler has been advocating, proving that Matt was telling us the truth when he said that he knows some CERN people who are equally fuzzy about the existence of the 125 GeV Higgs.

There is a lot of politics and lots of bizarre regulations that prevent these physicists from speaking. So I can't be sure whether Tuts is saying these insane things because he really has no clue about the actual odds or because he's forced to react in this way by some crazy speech codes gagging the ATLAS members. At any rate, what he says about the probability that the Higgs is around the 125 GeV bump is insane.

The collider has been running for 1.5 years and it has learned a lot. Most of the insights imply that the Standard Model's domain of validity is larger than many people were expecting before the LHC began its career. But from the panel discussion, you wouldn't learn a single bit. The most visible "positive" insight by the LHC so far is that there is a Higgs at 125 GeV. Now, the total combined evidence is just 4.7 sigma or so. (Both major detectors have collected 1/fb in 2012 so the total number of recorded collisions at this moment is almost certainly already enough to increase the combined ATLAS+CMS evidence to 5 sigma by now.)

And when a team collects 5 sigma, they will announce that they're certain and they will have discovered the Higgs. (The CMS and the ATLAS have probably already collected enough data by today so that they could publish a combined paper with a 5-sigma discovery.) But do you really think that the answer switches from "completely uncertain" to "certainly yes" overnight, as the confidence level jumps from 4.99 sigma to 5.01 sigma? You're joking, right? 4.7 sigma that exists today is almost the same thing as 5 sigma. It is extraordinarily unlikely that it is a statistical fluctuation, especially because it's sensible for the Higgs to be there. It's sensible to respect some rules that the official papers claiming a discovery build on 5-sigma excesses but that doesn't mean that a 4.7-sigma excess leaves the question completely open.

I would understand it if those people were hiding something totally shocking that makes billions of people move or panic. But billions of people don't give a damn about particle physics. If someone leaked the graph showing the observation of a totally new particle at the LHC, the number of people in the world who would want to see it – and who would try to spend at least 5 minutes by Internet searches to find the source – would be tiny: a hundred, a thousand? This whole secrecy seems to be detached from the reality.

In the panel discussion, experimenter Michael Tuts talked along with Brian Greene and two journalists, Dennis Overbye of the New York Times and Mariette DiChristina of SciAm. The journalists have of course pumped some degree of real interest and controversy into the discussion which was largely a positive contribution.

In my opinion, the work-in-progress – with all the disclaimers of its being faulty – should be much more available to the interested public and especially the scientific public. I don't really see what positive role the secrecy about the things that 7 billion people in the world are completely uninterested in plays in the scheme of things.

The discussion also touched the question whether "nothing new" or "something new" are equally good results. I agree with what's been said – that both of them are equally legitimate insights from the scientific viewpoint but the layman's (and not-so-layman's) emotional reactions to these two findings may differ and influence things such as the future funding. While the scarcity of new physics during the first year kind of agrees with some of my conservative expectations – one should find just "one bunch" of new physics every time he multiplies the energy by 10 etc. – of course that I would be more excited if the LHC were discovering lots of new particles and physical phenomena. And of course that there's no good reason to build a new $10 billion collider if you pretty much know it won't find anything new and important; under certain circumstances, the decision "not to build" would be totally rational.

Michael Tuts gave a good answer about the benefits of particle physics research – direct (for science) and good side effects. Brian Greene made a funny comment that not everyone has to be excited about physics; for example, after all those millions and other things that Brian has earned, his mom still thinks he should have become a doctor. So they argue whether Brian is a doctor or isn't a doctor today. ;-)

Tuts said that the reactions of neighbors in the airplanes to the proposition "I am doing particle physics" has changed from "oh" to "when will you find the Higgs" in recent years. He also gave an OK answer to the question whether physicists will ever give up the search for the Higgs. In 2012, it will decide about the fate of the SM Higgs in one way or another.

At 39:00, Brian Greene finally says that the Higgs mass is 125 GeV after a girl in the audience suggests, in a confused monologue, that the Higgs is massless. Brian says that the mass is 125 times the proton mass. Well, it's actually 133 times the proton mass: the proton is not as close to 1 GeV as Brian seems to think. ;-) But it's great that in a one-hour discussion on the Higgs, the visitors could finally – after 40 minutes – learn the most important previously unknown number describing the particle, namely its mass. He also tried to explain that the Higgs mass gives the mass to itself as well, by self-interactions.

An Indian gay asked a complex long question about the way how our theories are right and I had no clue what his point was. Brian said that physics uses mathematics and verifies theories. A black physics student asked a more meaningful question whether the Higgs discovery will illuminate the weakness of gravity i.e. whether it will get connected with more ambitious theories of physics. Brian says that it depends whom you ask. He believes that the claims that SUGRA or string models can predict the Higgs mass depend on lots of assumptions which means that he finds the support for such claims weak.

He is overstating the case of uncertainty here. There are various good reasons to believe that the low-energy effective theory is only the MSSM or MSSM with some particular extra charged particles or MSSM with some largely decoupled sectors. The probability that the right low-energy model is the "minimal one" is surely less than 100% but it doesn't mean that it's very small. There are very good reasons – such as the gauge coupling unification and the absence of new particles observed at the LHC – that the minimal SM/MSSM-like model (with some possible extra particles of a "very different kind") could be the right one.

Even if you consider additional possibilities, the number of possibilities is not one but it is not huge, either. Many models are instantly excluded because they predict unacceptable things. Models with too many colored things are gone because they screw the asymptotic freedom and other things. Once we adopt one of the viable possibilities, it becomes possible to predict particle masses as long as we demand some reasonable behavior near the fundamental scale. Those things are not guaranteed to be true but they may be true. It's totally plausible that the Gordon Kane et al. scenario is right. Brian's suggestion that those things are random guesses is just totally wrong.

To justify his totally fuzzy attitude of "nothing can be predicted", Brian Greene talks about those 15 orders of magnitude between the LHC scale and the fundamental scale. But Brian's opinion is a complete misconception. The existence of this gap does not mean that one loses the ability to predict low-energy parameters. The parameters are "pretty much the same ones" as those at the fundamental scale; they are just "running". Now, if they're running slowly – logarithmically – their values at the high scale and the low scale are close enough and even the relatively small difference may be calculated. If some parameters run quickly, as a power law, it means that their effects at low energies typically disappear completely if the effects were finite at the high scale. In both cases, we may use properties of the high-energy theory to make valid statements about the low-energy physics. Even if we assume something else than a big desert, we can make certain predictions, at least approximate ones.

At 46:00 or so, Brian Greene says it's tough being a string theorist. It may also be tough to be a phenomenologist if theorists on panel discussions spread the meme that phenomenology is just a guesswork. Dennis Overbye asked Brian what the particular value of the mass, 125 GeV, means for model building such as SUSY. Disappointingly enough, Brian has no clue about these matters, almost no clue about phenomenology. There are so many basic things about what this means for MSSM vs non-minimal models, the kind of mediation, the stop mass, or the stop mixing, but Brian hasn't even suggested that there is something here to learn.

But what Brian said around 48:30 made me really upset. He says that if the Higgs mass had turned out to be 142 GeV, they – or perhaps someone else – would write a paper analogous to Gordy's paper that would say that it had to be 142 GeV. Great, Brian, but the search for the right "analogous paper" that just happens to agree with the observations is the very purpose of science, not something one should try to humiliate.

The purpose of doing these experiments and measuring the Higgs mass and other things is to find out who is right and who is wrong. It's also true that there aren't any SUSY models in the literature whose lightest Higgs is 142 GeV. But even if there were such models, they would be legitimate. As we're collecting new experimental data, we're picking winners and losers, we're focusing on more viable and accurate theories surrounding the winners, and that's what the progress is all about. If Gordy's scenario is right, and a 125 GeV Higgs is a circumstantial evidence – while not a definitive proof – that it could be right, it's a great victory for science and for Gordy et al. and other people who helped to bring us to this state.

What would be the purpose of a speculation about another world in which the mass is 142 GeV and Gordy isn't a winner? I am probably not getting what Brian wants to say but it looks to me that he doesn't give a damn about the truth concerning cutting-edge phenomenology. He lives in those parallel worlds where every truth is a lie in a parallel world and he doesn't care much which of the universes is ours. I do care about it a lot. Whether a model predicts a right Higgs mass or a totally wrong mass is an important question and should strongly influence the fate of the model – and, to some extent, its authors.

When Brian was finished with this attack on Gordy Kane, Dennis Overbye said that he wasn't talking about Gordy at all; he was referring to Nima's statement that 125 GeV is a torture for theorists. This is of course an extremely oversimplified interpretation of what Nima thinks about the impact of the mass 125 GeV on particle physics. He has said many times that this mass makes sense, favors elementary Higgs, pretty much eliminates compositeness and technicolor, and even favors some classes of SUSY models. Some of his ideas could be associated with the word "torture" but one must be extremely careful not to misinterpret what the word "torture" was really meant to say.

A girl wanted to know whether the Higgs boson discovery would change the ideal gas equations or the way she is learning in pre-med. Well, no. Another guy asked how the LHC folks can ever be sure that the Higgs can't be there. Tuts explained that the Standard Model equations do predict how many Higgses should be seen given a particular mass. Brian Greene added a comment that is misleading, to say the least, that there are theories modifying the Higgs mass and couplings that would make it harder to see it at the LHC. There aren't sufficiently plausible theories of this kind. Brian's answer was of the type "anything goes" and pretty much reverted Tuts' important comment that qualitative assumptions behind your theories always have some quantitative consequences. One can't freely adjust all the couplings and think that it's still equally OK. In this part of the discussion and others, I got the feeling that Brian is actually one of the people who help to reinforce the breathtakingly stupid laymen's delusion that high-energy physics isn't testable, science, or whatever. Around 55:00 or so, Brian fixed it and said pretty much exactly what I did. But the term "fiddling" is still misleading. Good papers and sensible hypotheses – even those that are later falsified – are simply not about "fiddling", pretty much by definition.

At the end, the SciAm lady didn't say much interesting. Dennis Overbye said that a few years ago, particle physics was banned in the New York Times because nothing was happening. Now it's all over Gawker. Brian Greene said that it's exciting to try to link the physical of the small with the physics of the big – except that he previously said that he wouldn't be interested in any paper about that because papers have to make assumptions.

No comments:

Post a Comment