Thursday, October 29, 2009

Nature, NYT report the demise of Lorentz-violating theories

In August 2009, we discussed the preprint by the GBM/LAT collaborations working for Fermi, formerly known as GLAST:
Fermi kills all Lorentz violating theories,
which has ruled out all existing non-stringy theories of quantum gravity by confirming the rules of special relativity at a huge, trans-Planckian accuracy - as long as their parameters are chosen or estimated naturally.

Now, you may say that physicists know 5 or 12 or 2009 alternatives to string/M-theory - except that 4 or 11 or 2008 of them already reside at the dumping ground of physics.

Despite the paper's having 204 authors and despite hundreds of millions of dollars that the taxpayers pay for these great Fermi experiments, the important result was pretty much ignored by the news outlets (unlike the "results" of various surfer dudes that are promoted in all the media within a few hours). It simply wasn't convenient for the media that have been sleeping with the crackpots advocating all kinds of fundamentally misguided - and now dead - theories of physics for years.

The situation has just improved at least infinitesimally because the preprint was published in Nature:
A limit on the variation of the speed of light arising from quantum gravity effects
The popular article in Nature that summarizes is called
An intergalactic race in space and time
These texts in Nature are enough for some other semi-popular media outlets to take notice. The most catchy - and fair - title was chosen by Ars Technica:
Quantum gravity theories wiped out by a gamma ray burst
They could have also called it "Smolin's pants on fire". ;-) Physics World chose a less combative but equally fair title:
Special relativity passes key test
A more poetic and detailed version of the same title was also chosen by Dennis Overbye in The New York Times (whom I still consider the #1 physics journalist in the mainstream media):
7.3 billion years later, Einstein’s theory prevails
And Universe Today has played the same theme:
Einstein still rules, says Fermi telescope team,
much like PhysOrg.com and/or Science Daily:
Gamma-ray photon race ends in dead heat; Einstein wins this round,
Slashdot.ORG:
Intergalactic race shows that Einstein still rules,
and Symmetry Magazine:
Gamma-ray burst restricts ways to beat Einstein’s relativity.
They start by saying the essential point: sometimes, a single photon can tell us a lot.

Science News chose another good title with a very different, more microscopic focus (similarly to Ars Technica I started with):
Gamma-ray observations shrink known grain size of spacetime
Smaller dots, please. ;-) NASA/Fermi's own press release is also microscopic but doesn't reveal the punch line in advance:
Fermi telescope caps first year with glimpse of space-time
These summaries are pretty much fine although the far-reaching consequences are still being underestimated a bit. I think that the worst article - and especially title - about the story was chosen by Nude Socialist:
Universe's quantum 'speed bumps' no obstacle for light
This title clearly wants you to believe that particles different than the photon would see the hypothetical "speed bumps" i.e. violations of the relativistic dispersion relations.

Well, there is nothing special about the light (whose propagation depends on many other particle species running in the loops) in comparison with the gravitons, electrons, W-bosons, or other particles which means that no other particles would be annoyed by the "speed bumps", either: these "speed bumps" just don't exist. Ms Rachel was unfortunately misled by "zombie scientist" Mr Smolin, played by Ms Leslie Winkle in the famous TBBT scene below (I am Sheldon!).

OK, fine: Sheldon may also be Witten: go to 1:44:00 of this panel discussion at Strings 2005 (Real Video; TRF context) where the same TBBT exchange occurs between Witten and "singularly unconvincing" Smolin. Thanks to Rian. :-)

Mr Giovanni Amelino-Camelia, another "zombie scientist" whose every single paper (and the whole career) is about these dead Lorentz-breaking theories, claims that people like him have "fewer stomach aches now". Well, maybe they should visit a doctor if their stomachs work in this way. The very concept of Ms Rachel's selection of the "freshly falsified" scientists as the heroes of her article is just bizarre.

(Truth to be said, Overbye allows crackpot Smolin to spread fog, too.)

With one possible exception we just mentioned, the popular summaries mostly seem OK but it's still more important to look at the paper in Nature which has been slightly modified since the preprint version.

Peer review in Nature: a reconstruction

You can see the detailed effects of peer review on these changes. It's pretty much obvious that there have been referees from several groups.

First, it's very clear from the wording that there have been sensible referees - theorists - who actually understand physics and who have improved the original formulations by the experimenters. For example, the coefficient of the energy-dependence is usually expressed as an inverse mass scale. And this scale used to be called a "quantum gravity mass scale". You may check that the last sentence of the preprint's abstract said:
Even more importantly, this photon sets limits on a possible linear energy dependence of the propagation speed of photons (Lorentz-invariance violation) requiring for the first time a quantum-gravity mass scale significantly above the Planck mass.
That's a very bizarre term because quantum gravity doesn't imply any Lorentz violation (quite on the contrary: by gravity, we really mean "general relativity" which has the exact local Lorentz symmetry incorporated in its very postulates: these comments are irrelevant if you accept crackpots' monopoly to define the term "quantum gravity"), so the coefficients parameterizing Lorentz-violating terms shouldn't be called "inverse quantum gravity scales".

Indeed, this bug - caused by the experimenters' being linguistically brainwashed by the community of Lorentz-violating theorists - has been fixed. The abstract of the article in Nature explains the situation very well and uses the accurate, self-explaining term "scale of a linear energy dependence" for what used to be called the "quantum gravity scale".

On the other hand, I also feel that several of the "zombie theorists" who have been promoting these unmotivated theories for years and who still can't get used to the fact that their pet theories have been eliminated have influenced the article in Nature - in a negative direction -, too. That's a counterproductive effect that peer review leads to, too.

Although the natural estimate for the "scale of a linear energy dependence" that follows from the experiment is close to 102 times the Planck energy, it is being discussed only marginally and described as "not every secure". Well, the inequality saying that the Lorentz-violating mass scale is above 102 Planck energies is "not very secure", indeed. It is surely not a five-sigma statement. On the other hand, 102 times the Planck energy is the right order-of-magnitude estimate of the scale that follows from the Fermi experiment's analysis of GRB 090510.

And needless to say, the formulations that the paper has falsified the Lorentz-violating theories have been softened, too. The final portion of the Nature abstract claims the following:
We find no evidence for the violation of Lorentz invariance, and place a lower limit of 1.2 E_{Planck} on the scale of a linear energy dependence (or an inverse wavelength dependence), subject to reasonable assumptions about the emission (equivalently we have an upper limit of l_{Planck}/1.2 on the length scale of the effect). Our results disfavour quantum-gravity theories [3,6,7] in which the quantum nature of space–time on a very small scale linearly alters the speed of light.
It is a fair, cautious formulation, typical for careful scientists. When you are a physical theory who is dying (or who recently died) because the evidence has just falsified you, physicists diplomatically say that you are being disfavored.

Nevertheless, despite having no real complaints against the formulation, I would still guess that the referee(s) who forced the Fermi experimenters to weaken the statement in this way has (or have) either been one (or more) of the "zombie theorists", or a member of a competing experimental team that still wants to contribute to "new limits on Lorentz invariance" in the future. Some of the "zombie theorists" pushing the paper in this direction should have been made shut up by the paper instead of influencing it - but sociologically, it's very hard to design the mechanisms of peer review that would guarantee that it is the case but that would also allow other theorists to protect their (correct) theories against (wrong) experimental papers.

As you can see, I am convinced that a lot of personal interests play a measurable role in the behavior of the referees - and the peer-review process is surely not perfect because it is done by the people who are imperfect animals. Still, it is important or useful to eliminate or improve bad papers, at least statistically.

Because a substantial portion of the Lorentz-violating experimenters is working for the Fermi collaborations, you can't expect that the "remaining" experts in the world will have too much to say that the Fermi's own people couldn't see. This "monopoly" is both a good and a bad thing, depending on the context. It's good because the preprint was good to start with, because of the numerous experts in the teams, but it is bad because the "external protection against groupthink" is reduced.

I want to summarize this discussion by saying that we're surely not guaranteed that every choice or correction made by the referees in the peer review process makes the printed papers more accurate and more true. But the peer review still improves the quality of the printed papers statistically, unless the community that studies the discipline is genuinely corrupt.

And that's a moderately optimistic memo.

But let me also repeat that I consider this question - of the Lorentz invariance - settled and I don't think that any serious scientist, organizer, or sponsor of science should spend a significant amount of money, time, or effort on Lorentz-violating theories or the people who can't do anything else. And I even think that all future experimental projects that are primarily motivated by further tests of Lorentz invariance should be abolished, too - because all sensible doubts about the Lorentz symmetry have evaporated.

The situation is now analogous to the conservation of energy in mechanics, among many similar examples: we no longer test it at an "ever increasing accuracy" because we have both theoretical and experimental reasons to know that it's silly to violate the law.