The MAGIC gamma-ray telescope team has just released an eye-popping preprint (following up earlier work) describing a search for an observational hint of quantum gravity. What they've seen is that higher-energy gamma rays from an extragalactic flare arrive later than lower-energy ones. Is this because they travel through space a little bit slower, contrary to one of the postulates underlying Einstein's special theory of relativity -- namely, that radiation travels through the vacuum at the same speed no matter what?

The team studied two gamma-ray flares in mid-2005 from the black hole at the heart of the galaxy Markarian 501. They compared gammas in two energy ranges, from 1.2 to 10 tera-electron-volts (TeV) and from 0.25 to 0.6 TeV. The first group arrived on Earth four minutes later than the second. One team member, physicist John Ellis of CERN, says: "The significance of the time lag is above 95%, and the magnitude of the effect is beyond the sensitivity of previous experiments."

Either the high-energy gammas were released later (because of how they were generated) or they propagated more slowly. The team ruled out the most obvious conventional effect, but will have to do more to prove that new physics is at work -- this is one of those "extraordinary claims require extraordinary evidence" situations. But if the high-energy gammas really did lose the cosmic race, we're talking Big Discovery. It could be a way to constrain string theory, loop quantum gravity, and other bleeding-edge theories.

The basic picture is that high energies might cause small-scale fluctuations in the shape of spacetime, which would act as subatomic lenses. The higher the photon energy, the more it might induce such lensing and the slower it would cover large distances. Four minutes isn't much of a delay over a half-billion-year journey, but then again, you don't expect much. From the lag, you can deduce where quantum gravity kicks in. Some theories predict the effect is proportional to the quantum-gravity scale, in which case it happens at 5 x 1017 giga-electron-volts (GeV). In others, it's proportional to the square of the scale, in which case the lag implies 6 x 1010 GeV.

I need to look into this a bit more, but I just wanted to get the news out there for people to mull.

Update (August 23rd): Another co-author, string theorist Dimitri Nanopoulos of Texas A&M, writes to me: "I am very excited about this, because as you know we suggested this effect about ten years ago and we have follow through with several analyses and/or improvement on theory. Notice that the 0.4 x 1018 GeV is the typical string scale!!!!"

Daniel Ferenc of U.C. Davis on the MAGIC team writes: "There have been attempts to observe time lags in gamma flares and in gamma-ray bursts, but we have never seen something like this.... We should keep in mind that the effect may still be inherent to the process of the emission of gamma rays in the source, although not very likely. We are rapidly learning about such emission processes in AGNs from new data collected by MAGIC, HESS, VERITAS, and CANGAROO, in coincidence with x-ray and optical measurements, and will know more soon."

Update (August 24th): We're starting to see bloggers weigh in, including the inimitable Lubos Motl and Chris Lee at Ars Technica, though I'm surprised there's not more. Here we finally get some observations that probe string theory, if only tentatively, and people who have been loudly complaining about the lack of such observations have gone silent.

Update (August 25th): Peter Woit has now weighed in, though he has less to say about the MAGIC result itself than about a Slashdot headline about its being used to test string theory. I think Woit's comments miss the point somewhat. Like Samuel Johnson's walking dog, the fact we can talk about empirically probing quantum gravity at all is remarkable.

Update (August 27th): Backreaction is skeptical, and rightly so. Although caution is certainly in order, the whole reason that physicists (including Backreaction itself) are interested at all is that it might indicate a violation of special relativity.

Update (August 30th): Another co-author, string theorist Nick Mavromatos of King's College London, has this to say: "If the result is not a source effect -- and from only one observation of a single flare this cannot be inferred with certainty -- then indeed it would constitute a first positive detection of vacuum (subluminal) dispersion. In our paper we have argued against this effect being due to a conventional plasma effect, but as we say in the paper, still we cannot exclude other source effects that could result in delays of photons at the emission stage.... To be sure that the effect is a genuine quantum gravity effect, it has to observed also in all other instances ... especially in Gamma Ray bursts. There, the statistically significant population of bursts will be decisive for disentangling the effect from a source one."