Faster-Than-Light Neutrinos?

Probably not. But maybe! Or in other words: science as usual.

For the three of you reading this who haven’t yet heard about it, the OPERA experiment in Italy recently announced a genuinely surprising result. They create a beam of muon neutrinos at CERN in Geneva, point them under the Alps (through which they zip largely unimpeded, because that’s what neutrinos do), and then detect a few of them in the Gran Sasso underground laboratory 732 kilometers away. The whole thing is timed by stopwatch (or the modern high-tech version thereof, using GPS-synchronized clocks), and you solve for the velocity by dividing distance by time. And the answer they get is: just a teensy bit faster than the speed of light, by about a factor of 10-5. Here’s the technical paper, which already lists 20 links to blogs and news reports.

The things you need to know about this result are:

  • It’s enormously interesting if it’s right.
  • It’s probably not right.

By the latter point I don’t mean to impugn the abilities or honesty of the experimenters, who are by all accounts top-notch people trying to do something very difficult. It’s just a very difficult experiment, and given that the result is so completely contrary to our expectations, it’s much easier at this point to believe there is a hidden glitch than to take it at face value. All that would instantly change, of course, if it were independently verified by another experiment; at that point the gleeful jumping up and down will justifiably commence.

This isn’t one of those annoying “three-sigma” results that sits at the tantalizing boundary of statistical significance. The OPERA folks are claiming a six-sigma deviation from the speed of light. But that doesn’t mean it’s overwhelmingly likely that the result is real; it just means it’s overwhelmingly unlikely that the result is simply a statistical fluctuation. There is another looming source of possible error: a “systematic effect,” i.e. some unknown miscalibration somewhere in the experiment or analysis pipeline. (If you are measuring something incorrectly, it doesn’t matter that you measure it very carefully.) In particular, the mismatch between the expected and observed timing amounts to tens of nanoseconds; but any individual “event” takes the form of a pulse that is spread out over thousands of nanoseconds. Extracting the signal is a matter of using statistics over many such events — a tricky business.

The experimenters and their colleagues at other experiments know this perfectly well, of course. As Adrian Cho reports in Science, OPERA’s spokesperson Antonio Ereditato is quick to deny that they have overturned Einstein. “I would never say that… We are forced to say something. We could not sweep it under the carpet because that would be dishonest.” Now there’s a careful and honest scientist for you, I wish we were all so precise and candid. Cho also quotes Chang Kee Jung, a physicist not on the experiment, as saying, “I wouldn’t bet my wife and kids [that the result will go away] because they’d get mad. But I’d bet my house.” A careful and honest husband and father.

Scientists do difficult experiments all the time, of course, and yet we believe their results. That’s simply because it’s proper to be extra skeptical when the results fly in the face of our expectations: extraordinary claims require extraordinary evidence, as someone once paraphrased Bayes’s Theorem. When the supernova results in 1998 suggested that the universe is accelerating, most cosmologists hopped on board fairly quickly, both because we had a simple theoretical model in hand (the cosmological constant) and because the result helped explain several other nagging observational problems (such as the age of the universe). Here that’s not quite true, although we should at least mention that Fermilab’s MINOS experiment also saw evidence for faster-than-light neutrinos, albeit at a woefully insignificant level. More relevant is the fact that we have completely independent indications that neutrinos do travel at the speed of light, from Supernova 1987A. If the OPERA results are naively taken at face value, the SN 87A should have arrived a couple of years before we saw the explosion using good old-fashioned photons. But perhaps we should resist being naive; the SN 87A events were electron neutrinos, not muon neutrinos, and they were at substantially lower energies. If neutrinos do violate the light barrier, it’s completely consistent to imagine that they do so in an energy-dependent way, so the comparison is subtle.

Which brings up a crucial point: if this result is true (which is always a possibility), it is much more surprising than the acceleration of the universe, but it’s not as if we don’t already have ways to explain it. The most straightforward idea is to violate Lorentz invariance, a strategy of which I’m quite personally fond (although I’ve never applied the idea to neutrino physics). Lorentz invariance says that everyone measures the speed of light to be the same; if you violate it, it’s easy enough to imagine that someone (like, say, a neutrino) measures something different. Once you buy into that idea, neutrinos are an interesting place to apply the idea, since our constraints on their properties are relatively weak. It’s an interesting enough topic that there are review articles, and even a Wikipedia page on the idea.

And there are more way-out possibilities. Graininess in spacetime from quantum gravity might affect the propagation of nearly-massless particles; extra dimensions might provide a shortcut through space. This experimental result will probably give a boost to theorists thinking about these kinds of things, as well it should — there’s nothing disreputable about trying to come up with models that fit new data. But it’s still a long shot at this time. I hate to keep saying it over and over in this era of tantalizing-but-not-yet-definitive experimental results, but: stay tuned.

A few of the countless good blog posts on this topic:

This entry was posted in arxiv, Science. Bookmark the permalink.

95 Responses to Faster-Than-Light Neutrinos?

  1. Jan de Wit says:

    Were people even looking for electron neutrinos ‘a couple of years’ before SN 87A?

  2. Jim Cliborn says:

    Good blog! Keep us informed as understanding progresses. You are a good, clear thinker and writer, don’t give us up!!

  3. Lab Lemming says:

    Are there any astrophysical sources for this kind of neutrino, or are we stuck with terrestrial experiments? Skimming the papers suggests that aiming this thing at a detector in the pacific or antarctica would be very difficult.

    Anyway, my geologic interpretation is that they have the world’s most expensive earthquake detector.

  4. Pingback: Yatima » Blog Archive » nerdcore parenting

  5. Pingback: FTL Neutrinos? Not so fast…. « Whiskey…Tango…Foxtrot?

  6. Ellipsis says:

    very small correction — most, if not all, of the 24 neutrinos that were detected from SN1987A were electron antineutrinos, not electron neutrinos. (It is possible that 3 or 4 of the 11 events detected in the Super-K detector in Japan from SN1987A were neutrinos rather than antineutrinos, but this was far from clear.)

    (and yes, if neutrinos are Majorana, electron neutrinos and electron antineutrinos are just different helicity states of the same particle, but not if they are pure Dirac, etc etc etc. And I would personally bet about 1000:1 that the OPERA result is due to a systematic effect [but not 10000:1].)

  7. Jorge Laris says:

    Or maybe Speed light is increasing over time… Ok is very improbable.

  8. Raimo Kangasniemi says:

    I have to a bit sceptical about our ability to detect neutrinos from the 1987 supernova with the 1983 equipment. The amount of neutrinos caught in 1987 wasn’t terribly high even then.

  9. MrCompletely says:

    a nicely nuanced and non-dismissive interpretation of a nicely nuanced and non-hyperbolic announcement that has, predictably but unfortunately, resulted in a comically un-nuanced (perhaps even anti-nuanced) avalanche of headlines. it’s nice that we don’t have to dig too far for sobriety.

  10. Pingback: Bowerbird #22: Black Tie Gourmet Society « avian architext

  11. Mike Martin says:

    So, the only conclusion here is that faster than light neutrinos are expected but not probably? What about the implications with the mass of the Neutrinos. Can we call them tacchions?

  12. Brian Lacki says:

    Lab Lemming,

    There probably are, but we haven’t detected them yet.

    The way to make astrophysical neutrinos is basically to take a high energy proton or neutron and let it crash into something — either another nucleus or into a photon. If the collision is high enough energy (more than a few hundred MeV in the center of momentum frame), pions (and more exotic mesons) will be created. Those decay into muon and electron neutrinos and other particles, and muon neutrinos are what this experiment were using. (Also, neutrino oscillations I think will convert other neutrinos into muon neutrinos on astrophysical scales.)

    The problem is that neutrinos are also produced by cosmic rays hitting the upper atmosphere. These atmospheric neutrinos drown out astrophysical neutrino at GeV scales, at which this experiment worked. It’s only at high energies (TeV to PeV or higher) that you can look for astrophysical neutrinos with IceCube and other neutrino telescopes. If we think the speed of neutrinos increases with energy, then timing TeV neutrinos will be a more powerful experiment than timing GeV neutrinos.

    Of course, with this kind of experiment, you also want to do timing: a steady source of TeV-PeV neutrinos isn’t enough. Gamma-ray bursts might be sources of high energy neutrinos, and since they don’t last long, you could time the difference between when the neutrinos and the photons arrive. They’re also very far away, billions of light years, so any difference in speed will have a long time to accumulate into a large delay. The only problem may be if the delay becomes too long — if the neutrinos arrive years and years before the photons.

    But at the moment, we don’t know that gamma-ray bursts (or any other transient source) emit high energy neutrinos for sure: the Sun and Supernova 1987A are the only extraterrestrial sources of neutrinos detected so far.

  13. Doug says:

    Re #8: We did detect those neutrinos at a time when neutrinos moving close to the speed of light would have been expected to arrive from SN 1987A. Even if 1983 detectors wouldn’t have been able to detect it, you would still have to explain why an unusual neutrino burst consistent with a supernova explosion was detected at just the right time to fool us into thinking GR is correct for this supernova.

  14. Pingback: Misc § Unqualified Offerings

  15. Gene says:

    So if its true we have to give up Lorentz invariacne or causality? … I hate it either way.

  16. Pingback: Affaires insolites du 23 sept. 2011 : Des particules iraient plus vite que la lumière. Einstein va se retourner dans sa tombe. Einstien se serait-il trompé ? Maï Salaün « Maï Savoir Is Riche

  17. Rick says:

    “So if its true we have to give up Lorentz invariance or causality? … I hate it either way.”

    I don’t think we’ll have to give up either when this is sorted out. But, if we have to, I vote for giving up Lorentz invariance.

    In the meantime, this is the best approach: http://www.xkcd.com/955/

  18. Gavin Flower says:

    Has anyone considered that going through matter, might cause these neutrinos to go a bit faster than light? If that is correct, then comparing this experiment to neutrinos travelling through space may not be valid!

  19. Pingback: And even more on super luminal neutrinos | The Zingularity

  20. jimthompson says:

    I’m rather disappointed with the response of theorists to this paper: isn’t it a theorist’s JOB to immediately start writing papers and talking about all the implications if this result IS correct? Not a peep , that I’ve seen, about what OTHER experiments might or might not see etc. etc. The fact that the paper is probably wrong has NEVER seemed to stop theorists in the past……After all, this paper is probably NOT going to be refuted for at least a week, perhaps much longer.

  21. Ranger Dan says:

    Did they try putting new batteries in their calculator?

  22. Torbjörn Larsson, OM says:

    So … has everybody caught where they goofed yet?*

    It is an easy one. According to the paper the distance measurement procedure use the geodetic distance in the ETRF2000 (ITRF2000) system as given by some standard routine. The european GPS ITRF2000 system is used for geodesy, navigation, et cetera and is conveniently based on the geode.

    I get the difference between measuring distance along an Earth radius perfect sphere (roughly the geode) and measuring the distance of travel, for neutrinos the chord through the Earth, as 22 m over 730 km. A near light speed beam would appear to arrive ~ 60 ns early, give or take.

    Of course, they have had a whole team on this for 2 years, so it is unlikely they goofed. But it is at least possible. I read the paper, and I don’t see the explicit conversion between the geodesic distance and the travel distance anywhere.

    Unfortunately the technical details of the system and the routine used to give distance from position is too much to check this quickly. But the difference is a curious coincidence with the discrepancy against well established relativity.

    ——————–
    * Extraordinary claims need extraordinary evidence. Other outstanding concerns are:

    1. This needs to be repeated.

    2. It is not a clear photon vs neutrino race. Physicist Ellis and others here noted that the time differential for the supernova SN 1987A was a few hours, but at the distance of ~ 200 000 ly it should have been years if the suggested hypothesis would be correct.

    3. Analogous to the experiments where light waves seemingly travels faster than photon speed in vacuum, they don’t measure travel times of individual neutrinos but averages over a signal envelope. That must be carefully measured to establish that particles (or information, for that matter) travels faster than relativity allows.

    Especially since the neutrino beam oscillates between different kinds of particles!

  23. Isn’t it true that in this context “giving up Lorentz invariance” means just for a few exceptions like muon neutrinos? As I understand it, current quantum physics is based on Special Relativity and there’s a lot of experimental evidence for that basis using many different particles of different masses. Anyone have a theory as to why most particles would be Lorentz invariant, and only one or few would not? And especially why one form of neutrino is and another is not (based on the Supernova 1987A observations).

    Of course there was a similar choice awhile back, when Bell showed (and Aspect and other experimenters proved) we’d have to give up either locality or causality, and most chose to give up locality.

  24. JK Finn says:

    From the arxiv paper ( http://arxiv.org/abs/1109.4897 ):

    In terms of interactions in the detector, the νμ contamination is
    2.1%, while νe and νe contaminations are together smaller than 1%.

    Wasn’t there some interesting result last year about a possibly significant difference between the masses of νμ and νμ? If true, wouldn’t this be likely to distort the distribution of arrival detection enough to shift the calculated value?

    Still reading the paper, maybe it is mentioned there…

  25. Paul says:

    jimthompson: actually, theorists have been writing papers about tachyonic neutrinos for years. There was some motivation from tritium decay experiments, where some data can be interpreted as the electron antineutrino having a slightly negative mass^2.