Of Particular Significance

Neutrinos and Light: Summary and an Open Space for Questions

POSTED BY Matt Strassler

POSTED BY Matt Strassler

ON 09/26/2011

As befits our information age, and anything that has to do with Einstein, we are now deluged with a blizzard of detail, commentary, speculation, polemic, and downright silliness surrounding the recent OPERA experiment (the one which claims to find neutrinos arriving, after a trip from Western Switzerland to Central Italy, earlier than expected.) I aim to avoid adding to the confusion, and will only post what I feel is both clear and reliable.  Right now I have little new to say that fully satisfies both criteria.  However, later in this post you will find a summary of the posts I’ve put up so far, to help you navigate what’s already here.

One goal of this site is to create a space where lay people with a range of backgrounds can ask questions. While I deeply appreciate that many of my physics colleagues are reading my posts and some are commenting on them — along with highly educated non-physicists who’ve been asking very sensible questions and making insightful comments — my one concern right now is that the level of comments is so sophisticated that a person less familiar with the physics will feel intimidated about asking more elementary questions.  So I’d like to ask that the comments on this post be limited to those who feel there are some very, very basic points about the OPERA experiment, and the relevant background, that they just can’t follow. 

I’m sure I won’t have time to answer all of your questions, at least not right away — and sometimes an answer won’t be possible without an extension of the website. But just knowing what your questions are will help me decide what material to add to this website, to make it more useful for the more general reader.  So: Please Ask Away. [Experts: please see additional comment below]

Finally: the summary of the posts so far.

If you find these posts are too advanced for you or leave a burning question unanswered, feel free to ask questions in the comment section below. Again, I can’t promise an answer, but will try to answer as many as possible.

A last comment for experts: if you see a question from a layperson sitting here, and you feel you can answer it, I certainly won’t stop you. But I am afraid I must reserve the right to delete the answer or to edit it if I feel it isn’t both correct and crystal clear, as well as free of jargon to the extent possible. So please, if you give an answer, try to make sure you are very comfortable with its form, that you’ve proofread it very carefully for clarity, style and content, and that if possible you’ve tried it out (now or previously) on a non-expert. Obviously this is a tricky business as one can answer any question at multiple levels; just please make your answer something that a substantial segment of the public could follow.  [And if you see what you think is a mistake in one of my answers, please let me know.]

Share via:


86 Responses

  1. The reason light takes longer it because it bends around mass. Understanding the difference may present solutions to interstellar travel and neutrino conversion.

    1. Interesting suggestion. Nice words. Equations count for more, however. I wonder how you explain why the neutrinos from supernova 1987a arrived within a few hours of the supernova’s light, despite the fact that, according to your thinking, the light had to follow the curved space between here and the supernova, while the neutrinos did not.

      1. Great comment. I lend more support to a shock-wave effect causing an anti-matter ripple that initially slow neutrino’s. I believe there is more work needed to understand the relationships. I feel the mass density of, and the distance/time from, the occurrence are significant contributors to the equations with regards to neutrino oscillation along the ripple. Also, something like gravitational pull from black holes have produced a slingshot effect and that may be another factor.

  2. Isn’t the 1987a super nova result (photons and neutrinos traveling at the same speed) problematic if neutrinos have mass?

    1. Not if the mass is small enough. You are right, the supernova neutrino observations put an upper bound on neutrino masses, because if the masses had been large enough, the neutrino speeds would have been far enough below light speed that the neutrino arrivals would have been noticeably spread out in time. But any neutrino with a mass-energy below a few 1000s of eV was ok, if I remember. Meanwhile it is typically concluded, from neutrino oscillations and cosmology, that neutrino masses are probably below 1 eV, with the heaviest only about .1 eV. The actual masses remain to be measured.

  3. The OPERA experiment tries to match a proton “probability distribution function” (W(t)) with the rather sparsely sampled neutrino “probability distribution function” (protons ~ 10^9/ns, neutrinos ~ 0.5/ns), by computing

    Product( W(ti + δt) where ti are the neutrino arrival times, and δt is chosen to maximize the product, and is interpreted as the delay.

    The problem with this measure is that it is significant only where (W’/W) is large, i.e., at the edges. My numerical experiments using a reduced data set measured from the edges of the first extraction in the OPERA paper show that just removing the single earliest neutrino in the first bin in the first extraction or adding a neutrino in the (empty) second 50ns bin in the first extraction is enough to shift δt by 5-6ns. The standard error really depends on what you have at the edges – where the statistics are the worst.

    My question is that is there not a better estimate for δt using the Kolmogorov-Smirnov test?

    The K-S test uses cumulative probability distributions (CPDF) and is less sensitive to the details of the edges than the OPERA measure. The K-S statistic is max( CPDF1-CPDF2) where CPDF1,2 are the two cumulative probability distributions. One would find the δt which minimizes the K-S statistic. If the K-S statistic turned out to depend on the beginning or tail of the CPDFs then we’d know the problem is with the poor statistics at the front and tail of the PDFs. But really, the K-S statistic should be arising from the middle of the CPDF where the accumulated statistics are good.

    Thanks in advance for your feedback!

  4. I don’t know of any convincing evidence that light at different energies travels at different speeds. There are many sudden processes in nature (such as Gamma Ray Bursts) that indicate that photons at different energies arrives at essentially the same time on earth. Yes, at any given time there might be a hint of something odd, but there are always hints of something odd and most of them go away over time.

    Generally, until I look at the math behind the words that describe a theory, there’s no way for me to know if the words make sense. Quantum mechanics, if you described it in words alone, would sound completely insane; it’s only the fact that it has precise and well-defined mathematics, and you can calculate the precise energies of atomic transitions, matching thousands of experiments, that convinces anyone that it is essentially right.

    As for the neutrinos — I no longer think OPERA could be correct, based on simple theoretical arguments and on existing data from other experiments. When I sure I fully understand the argument and any loopholes, I’ll post.

  5. Professor Strassler,

    If it is an energy effect, then, presumably, repeating the OPERA experiment with lower energy neutrinos would show that — e.g., at energies comparable to those of the neutrinos from the SN. I’m assuming that there are technical reasons for not having done that experiment. Can you say what the are and if such an experiment, if it makes sense, is on the cards?


    1. The really interesting experiment to do would be one at *higher* energies than OPERA, since then the effect would be larger and much easier to observe. Such an experiment, if it saw nothing, would indicate OPERA must be wrong. If the effect decreases with energy, then a measurement at lower energy would be unlikely to see anything — and so neither confirm nor contradict OPERA.

      1. Thank you for the reply. Following up on the energy issue — I can see the rationale in going to higher energy to see if the effect gets bigger, but that is assuming that it exists at all and that it does increase with energy. Given the actual Supernova observations and the suspicion that the result might be due to a systematic error, wouldn’t the lower energy experiment make sense as a calibration? If the 60 ns discrepancy persisted at the energy of the SN photons then it seems like that would be a clear indication of experimental error. Is it difficult to change the energy of the OPERA neutrinos?

        1. There are a lot of ways to go after this problem, and it isn’t obvious which is best. I’m now convinced it is an experimental problem, and it is just a matter of finding the mistake; the result almost certainly is wrong. Since that’s the case, the best way to do that would be to find OPERA’s mistake in situ, not build a new and expensive experiment whose only purpose was to check an almost certainly wrong result. But there are a lot of problems with lowering the energy of the OPERA beam. It took them three years to do this measurement. With a lower energy, the neutrino collision rate decreases, the experiment (designed to observe higher energy neutrinos) would become inefficient and eventually ineffective, and the challenges of the experiment would not really be addressed. So I suspect checking OPERA directly would be best done by changing the beam to make the pulses of neutrinos much sharper and shorter. However there are technical challenges that I do not know much about, so I don’t know if that’s actually possible.

          Perhaps someone can use an existing experiment, or design one with a secondary purpose. But still, the first priority would not be to go to low energy but to go either to the same or to higher energy, to see if the OPERA effect can be reproduced. Given what I know now, that seems pretty unlikely, so again this experiment ought to have a secondary purpose to justify the expense.

    1. No, I am afraid I cannot. The summary could not be brief, for one thing. Moreover, we do not know what these findings (which right now consist of a single data point) mean theoretically, so the implications cannot be known at this time. And finally, even if I knew what these findings meant theoretically, their implications would take considerable time to work out. Top quality science is rarely done in a week’s time.

  6. One more question.

    Are neutrinos supposed to emit any Cherenkov radiation while traveling faster than the local speed of light in a medium?

    1. They aren’t electrically charged, so they won’t emit photons in the usual way. However, Cohen and Glashow have apparently pointed out that they will still, through the weak nuclear force, emit electron-positron pairs (via effects of Z particles). I have to read that paper today.

  7. 1. How does the current experimental upper bound on the absolute magnitude of error term for the lack of dispersion of light in vacuum look like as a function of frequency?

    2. What is the upper bound on the absolute magnitude of the difference between average back-and-forth speed of light in vacuum along any two perpendicular directions (as determined by current state-of-the-art Michelson-Morley like or any other experiment)? Is 0.2 mm/sec above or below this limit?

  8. 2 ideas:
    * Could this apparent faster-than-light neutrino travel be related to the faster-than-light tunnelling which had some media coverage around 1994, in the sense that while particles / waves might statistically exceed speed of light it’s still impossible to actually transmit information faster than light (thus not even the theoretical possibility of exploiting this for time travelling)?
    * Could this phenomenon be related to not only energy but also that neutrinos have to travel through matter / through earth’s gravitational potential?

    1. I don’t see how this effect could be due to either of your suggestions. The experiment, if correct, *would* allow information to be transmitted faster than the speed of light-in-vacuum (in some sense, if OPERA is correct, it already was, because information about the proton pulses that made the neutrino pulses arrived earlier than it should.) The second suggestion should only have been able to slow neutrinos down, in Einstein’s theory.

  9. Hi Matt, brilliant blog by the way, I hope when you get back from London you can respond to this question. After watching the OPERA CERN talk and then skimming their paper I still cant understand how they know what the delay is between a neutrino arriving at their detector and the appearance of a photon in their scintillator (I think they test these with a laser so that bit of the delay is clear). If there was a delay of 60ns in the process of photon production from a neutrino then all the mystery is gone. Presumably they can’t test this with a local beam of neutrinos so how can they tie that figure down?

    1. Following up on the issue of experiment design, I was intrigued by John Butterworth’s comment in the Thursday, 29 Sep 11 Material World pod cast bit.ly/nO8aTH about the detector only picking up a tiny fraction of the solid angle of the wide neutrino fan produced in CERN. If I understood him correctly, he suggested that perhaps the neutrinos heading in the direction of the detector could have been generated relatively early in the CERN source pulse. Does that make sense to you?

      1. Butterworth is very smart and experienced, and I have no reason to think he’s wrong in a silly way. On the other hand, I’ve heard at least a dozen reasonable proposed explanations from comparably good experimentalists. So there’s no reason to think he’s right either. I’ll bet the OPERA people would argue that any such effect as he describes would have distorted the distribution of neutrino travel times that they observe, rather than just shifting it over by 60 nanoseconds.

        I think the theoretical arguments against the result are becoming a lot more serious.

  10. I think I am confused about the energies of the neutrinos. In various places I’ve read about the SN1987 neutrinos having one energy, and the ones produced at CERN having another. What kind of energy are we talking about here?

    It can’t be momentum, since we’re obviously not even sure how fast they are going. I assume it isn’t mass, since we don’t really know that either for neutrinos. I know that photons have an energy depending on their wavelength – is there a corresponding concept for neutrinos?

    1. The energy here is just ordinary energy: energy of motion plus mass-energy. This energy is related, by quantum mechanics, to frequency; they are proportional to one another. (it is energy and frequency which are related, and it is momentum and inverse-wavelength [i.e. 1 divided by wavelength] which are related; that is true for photons, neutrinos, electrons, W particles, and everything else. For photons and other massless particles, energy and momentum are proportional to one another, so your statement about photons’ energy and inverse-wavelength happens to be correct. It’s almost correct for particles whose motion-energy far exceeds their mass-energy, as for neutrinos with GeV energies, but one should be precise; energy is related to frequency, not inverse-wavelength.)

      We can determine neutrino energies in two ways. The first, when we make a neutrino beam (http://profmattstrassler.com/2011/09/23/how-to-make-a-neutrino-beam/), is to know the energies of the protons that we used to make the neutrinos in the first place (which we need to know anyway to make the proton beam pulses) and either calculate the energies of the neutrinos that emerge after the complicated process of making them, or infer the neutrino energies by measuring the energies of the anti-muons that are produced along with the neutrinos in pion decay. The second way to measure neutrino energies is when we detect them through their collisions with an atomic nucleus or electron (http://profmattstrassler.com/2011/09/25/how-to-detect-neutrinos/) . If we can measure the energy and momentum of the ensuing debris, then (remembering that the total energy and momentum are always the same before and after the collision) the colliding neutrino’s energy can be inferred, completely or roughly depending on the details of the collision.

  11. Sir,
    I have one query regarding “Twin Paradox”.

    Lets take the example of stationary A and moving B (lets say at .9999c).

    Lets say at some instant B sends message/tick to A, that he is 10 y/o as per B’s clock. A would get this message, lets say, when A is 20 y/o (as per A’s clock). and it should be same other way round too.

    Now at that instant (when at receives that message), would B also not have aged to 20? Even though A would not know about it.

    So, essentially both should age at same rate (from neutral perspective), and it should just be the time gap in knowledge of their ages with each other – as per their own individual clocks.

    So, as per Twin Paradox, where does the ‘difference in aging’ come from – when they are moving away from each other – near speed of light – relatively, but same for each other?

  12. Can you please explain this to me why do processes like the one depicted in this picture: http://imageshack.us/photo/my-images/97/feynmandiagram.png/
    not happen and then right after it the opposite process. Having such a process for a short while should lower the speed of a photon slightly but since it probably has a low change to happen and since the particle-anti-particle state will decay into a photon again fast it should only lower photon speeds slightly below the speed of light. Also i think this kind of process is compatible with the standard model.
    I would very much appreciate an answer .

    1. I cannot currently think of a way to explain the answer to this question without some technical discussion. I’ll try to find the time to put an answer to this question into the technical zone, or to come up in future with an answer that is both intuitive and correct. But suffice it to say that one can easily prove, using the mathematics of quantum field theory, that a massless photon simply cannot obtain a mass from quantum fluctuations in the vacuum (also known as particle-antiparticle fluctuations, also know as loop effects, also known as what you tried to explain above.) And in Einstein’s theory a massless particle moves at the speed of light. End of story. It may not be immediately obvious or intuitive, but it is correct.

      1. Thank you i think it suffices as explanation that what i tried to use is a loop-process hence only virtual particles are create hence no massgain even for short moments. So the diagram i tried to use is only half of a loop process the opposite process completing the loop. Did i understand that much right?

        1. Yes, you are right that, in terms of diagrams, what you tried to use is a loop process, or more colloquially, it is the interaction of a photon with the fluctuations in vacuum of the electron field. It still is not trivial, however, that a loop cannot give a contribution to the photon’s mass. For example, the same loop *will* give a correction to the mass of a spin-zero particle like the Higgs particle — and this fact bears a deep connection to the hierarchy problem, which you can read a bit about here: http://profmattstrassler.com/articles-and-posts/particle-physics-basics/the-hierarchy-problem/

      2. I now read a bit about the hierarchy problem am i right in my understanding that so far it is not yet clear if it is not just a problem arising from series expansion you do with feynman diagrams? So the problem might actually vanish if the fieldtheory would be treated without any estimates which is not possible analytically yet?

  13. Totally naive question here: Generally, neutrinos do not interact with the stuff they pass through. But you noted, in the context of photonic booms, that very-high-speed charged particles in water can travel faster than light. This suggests, that at least for some sorts of particles, the medium through which it travels makes a difference to their “speed limit.” If this were true for neutrinos too, might not a geological anomoly between the generator and detector at the test site cause the observed result (if those observations are confirmed)?
    Displaying more ignorance than I’m comfortable with,

    1. Again, you have to be very precise in language (this is key to being a good scientist.) In a medium where light travels slower than light travels in vacuum, other types of particles in the medium can travel faster than light travels in the medium, but slower than light travels in vacuum. It is inconsistent, however, for any particle to travel faster than light travels in vacuum, unless you modify Einstein’s relativity and with it the very geometry of space and time.

      Neutrinos traveling in rock could in principle travel slower than the speed of light in vacuum, due to some hitherto unknown interaction between neutrinos and matter. (That might be excluded by other experiments, but at least it’s legal.) But no such anomaly could cause neutrinos to travel faster in rock than light does in vacuum. That would be in violation of the mathematics of Einstein’s relativity, and require the theory of relativity to be modified. (More precisely, it would be possible within Einstein’s theory only if neutrinos were tachyons, but no sensible theory of tachyons has ever been proposed, to my knowledge.)

  14. Dear Matt,
    now they match the neutrino profile arriving in Gran Sasso to the proton profile which produced the pions at CERN. A lot of complicated things happen between the moment the protons are being produced by the SPS and the the moment the neutrino’s are finally created. Now the neutrino’s are produced together with the anti-muons. Could one not detect the latter at CERN and compare their profile to the Italian neutrino’s?

  15. 4-5 question from a layman (me):

    Is it proven that neutrinos does not have electric charge ?

    If they have no electric charge as detected so far, could this be due to the fact that they have so little mass/are so much smaller than the other elementary particles, that their electric charge is way to small to be detected ?

    And is it possible that the speed of light (c – speed limit) that is used in mainstream physics today only holds true for elementary-particles that have electric charge as detected so far, and, hence, that the speed of light (c – speed limit) in reality is slightly higher due to the neutrinos since they have so far a smaller and non-detectable electric charge ?

    And what if the speed limit C is slightly higher, due to the new result from OPERA, couldn’t we just use this new speed limit as the new C in the theory of special relativity ? Wouldn’t the theory of special relativity hold true then even tough the C was adjusted to the new speed limit ?

    1. I apologize for having proposed theese dumb questions. But I have now tried to answer some of them myself:

      Regarding my last question: “And what if the speed limit C is slightly higher, due to the new result from OPERA, couldn’t we just use this new speed limit as the new C in the theory of special relativity ? Wouldn’t the theory of special relativity hold true then even tough the C was adjusted to the new speed limit ? ”

      My answer is of course: No. If C is adjusted then the theory of special relativity will no longer predict exact the time muon lives when they are moving near light speed. Thats a fact proven by a number of experiments, if my understanding of this is correct.

      And regarding my question: “And is it possible that the speed of light (c – speed limit) that is used in mainstream physics today only holds true for elementary-particles that have electric charge as detected so far, and, hence, that the speed of light (c – speed limit) in reality is slightly higher due to the neutrinos since they have so far a smaller and non-detectable electric charge ? ”

      My answer is: C has to have the measured value we operate with in special relativity. Or else all mainstream physics will collapse, in my opinion.

      Regarding my two first questions: I understand if Professor Strassler don’t see the point of answering those questions since they may be meaningless questions.

      Best regards
      Stein Sivertsen

      1. They’re certainly not dumb questions. For question 1, I think you are right, but I think this is not such a strong constraint, and I think there are much stronger constraints from other measurements, such as the lack of Cerenkov radiation from high-energy muons. For question 2, the question is quantitative (how much of a difference in speeds could there be and still be consistent with experiment.) Indications are, from what I have read, that the constraints are again very strong. It seems impossible to have an effect as large as what OPERA measures and still be consistent with other measurements.

  16. (While I’m being a little lazy and not sifting through the good, the bad, and the ugly of Google results here is another question (questions).)
    1. a faster than light neutrino is a tachyon ?
    2. is it correct to say that tachyons / FLT neutrinos have negative mass?
    3. is negative mass repulsive against positive mass ?

    They say there is no such thing as a stupid question. I have my doubts about that. Any response to my questions would be gratefully recieved.

    1. 1. If we assume Einstein’s theory is correct, any particle traveling faster than light travels in a vacuum is by definition a tachyon. What is a tachyon? It is a particle traveling faster than the speed of light in the context of Einstein’s theory.

      2. No. Tachyons have a negative mass-squared. That means they have an imaginary mass, in the precise sense of “imaginary” used in the theory of “complex numbers”. You might wonder how that can be consistent with E = m c-squared. Remember that this equation holds only for particles at rest., and a tachyon always travels faster than light and can never be brought to rest. Energy is always positive or negative, never imaginary.

      3. Negative mass-squared is not repulsive against positive mass-squared. In Einstein’s theory of gravity, gravity pulls on energy (which is proportional to mass for very slow-moving objects like the earth relative to the sun, giving Newton’s law, but not for photons, which are massless but have energy and are still pulled on by gravity). Actually it is more complicated than that, but basically energy and momentum are the things that matter. And a tachyon’s energy is always positive.

      However, this is all probably moot. In quantum mechanics, it appears that tachyons lead to instabilities in the theory; in trying to make a tachyonic particle you would instead rearrange the whole universe. I have never seen it demonstrated that one can even make a sensible theory of a tachyon. So I would expect that were this measurement correct, it would mean Einstein’s theory of relativity must be modified, not that neutrinos are tachyons and that Einstein’s theory is correct.

    1. Yes — you should remember that rock is not very “solid”. An atom is mostly empty space, with a nucleus whose volume is .000,000,000,000,001 of the atom’s volume; what appears solid to us is basically mostly vacuum, plus some electromagnetic fields from the electrons and atomic nuclei that also don’t affect whether vacuum fluctuations occur. So although rock may have some effects of its own, it does not change the fact that a high-energy particle traveling through ordinary matter spends most of its time in vacuum, which is full of important quantum fluctuations.

  17. Excellent Site. Maybe IF the results of neutrion velocity are correct the question to ask is not why they are moving faster than light (photons) but why photons are travelling at a lower velocity. This may lead to light (photons) having a latent mass and weak interaction with some known or unknown fields. This may lead to some energies for matter above E equals mc squared or latent energy in addition to matter anti-matter annilation. Any speculations.?

  18. Can’t make a definitive statement yet on this. But … even if the paper (which isn’t stupid, but walks very close to the edge) made sense to begin with as an idea in theoretical physics — which I have considerable reason to doubt — I suspect this option is now excluded by data, as we do know vastly more about neutrinos than we did then, thanks to supernova 1987a and numerous direct experiments. For example, I think this option would have given something quite different as far as the supernova neutrino observations. But one of the authors, Kostelecky, is an expert on violations of Einstein’s version of relativity, and on weird effects involving neutrinos, so he might have a trick up his sleeve that I might not know about.

    Meanwhile, OPERA certainly appears inconsistent with this option, given the supernova neutrinos. A tachyon (if any could actually exist without being quantum mechanically inconsistent, which I highly doubt) would, according to the mathematics, travel faster as its energy is lowered. At zero energy, a tachyon would have infinite speed. [Sorry — that’s what the math says. You don’t have to believe in these things — yet.] Thus the supernova neutrinos, which are at lower energy than those at OPERA, would be expected to travel faster, were tachyons possible. But we observe supernova neutrinos traveling very close to the speed of light, while OPERA claims its more energetic neutrinos are arriving early. So the dependence on energy versus velocity of a tachyon is simply wrong. [And there is very strong evidence (from neutrino oscillation experiments, to be explained elsewhere) that neutrinos of different types travel at the same speed for a given energy, so the fact that supernovas and OPERA make neutrinos of different types should not matter.]

  19. This is an amateur question.
    Could someone explain … “Then suddenly there is a flurry of neutrinos, carrying much higher energies than usual… and then, back to the drizzle.”

    I’m assuming “kinetic energy”.
    What imparted that kinetic energy to what?
    I can understand bringing a ball at the top of a building increases the potential energy which gets released as kinetic energy when you drop it.

    1. Consider the chain of events: a) you hold the ball at the top of a building; b) then you let go; c) the ball falls; d) it smashes to bits and the pieces come to a stop due to friction; e) the low heat generated by the friction gradually radiates back into the air in the form of infrared (low-energy) light.

      Now consider this: a) a star is prevented from collapsing on itself due to the pressure from its internal furnace; b) the furnace gives out and c) the star’s core begins to collapse on itself; d) powerful destructive processes occur inside the collapsing core, including the conversion of electrons and protons to neutrons and neutrinos, and tremendous heat (random motion of particles) is created; e) that high heat is radiated into space in the form of high-energy neutrinos (and, to a much lesser extent, light.)

      Does the analogy make sense?

  20. First of all and to be original, congratulations for your blog and your effort in presenting last minute science to all of us layman
    My question is the following, I’m sorry in advance if I say a nonsense and also for the language so I’m not english speaker:
    It seems Supernova 1987a is a strong argument against FTL neutrinos, in order to beat it it is said that energetic neutrinos could travel faster than low energy ones. According to Einstein’s, mass and energy are equivalent and zero-mass particles travel at the speed of light, so energetic neutrinos should be heavier and therefore, would it be more difficult for them to travel faster than the lighter neutrinos like the ones of 1987SN whose v=c?


    1. Your question involves a profound and very common misconception: mass and energy are *not* equivalent, and energetic neutrinos are *not* heavier. You are not alone: many people incorrectly interpret E=mc2 to imply that as a particle’s energy increases, its mass increases too; but in fact the equation refers only to particles that are not moving. Freshman textbooks unfortunately contribute to this misconception! For particles that are moving, E is LARGER than mc2, and the mass m remains the same whether the particle is moving or not.

      I explained this error in more detail, through an interesting story, in a reply to a question in the comment section of the Higgs FAQ, http://profmattstrassler.com/articles-and-posts/the-higgs-particle/360-2/ . Search for “Alfa” (the person who asked the question) and you’ll see both the question and my answer. Then if you are still confused, please return here and re-ask your question.

      1. Thanks, I see my error now. I could now rewrite my question, I’m sorry if it’s silly again: saying that the reason neutrinos are faster than light could be because they are more energetic is not a kind of circular statement?

  21. A number of questions:

    a) For SN1987A; do we know roughly the spread of energies in the detected neutrinos? Some 20-odd events is not much, but if we have neutrinos spanning an order of magnitude in energy arriving within 13 seconds of each other, that should constrain the energy-dependent speed, therefore the difference to light speed, quite a lot, right?

    b) Does an oscillating neutrino travel at different speeds depending on its current rest mass? I assume the energy must be constant and the rest mass changes when it oscillates. How does this work for slow-moving neutrinos? For example some far redshifted cosmic neutino background ones. Do they stand almost still as near-100%-tau and speed up considerably when they’re near-100%-electron? How do they know in what direction to gain momentum when they lose mass?

    c) What are our chances of detecting neutrinos from another supernova “soon”? Has our ability to detect them grown enough since 1987 that we could detect one from the Andromeda galaxy, or are we limited to Milky Way supernovas for these purposes? How often do we get a Milky Way supernova? Do we have the ability to measure energy vs arrival time well enough to know “almost exactly” what the rest mass of the neutrinos are the next time we see a Milky Way supernova?

  22. first of all, thanks for this site – it’s terrific! here’s a question – i’ve read that neutrinos can change from one type to another, but that this can happen only if they have a rest mass – although it can be very very very small, it’s got to be there (right?) if so, how can a neutrino possibly approach light speed (like those, say, from the supernova a few years back) without its mass approaching infinity (according to relativity)?

    1. There are two layers to your question, but the first involves a profound and very common misconception. You are under the impression that as a particle’s energy increases, its mass increases too. You are not alone: many people incorrectly interpret E=mc2 to refer to particles in motion as well as at rest. Freshman textbooks contribute to this mistake. I explained this error, through a story, in a reply to a question in the comment section of the Higgs FAQ, http://profmattstrassler.com/articles-and-posts/the-higgs-particle/360-2/ . Search for “Alfa” (the person who asked the question) and you’ll see both the question and my answer. Then if you are still confused, please return here and re-ask your question.

  23. A question that I’ve been kind of afraid to ask elsewhere because a lot of people seem to think it’s not relevant to these lofty matters, but what would the actual practical implications be if there turned out not to be any systematic errors or theoretical flaws, and others were able to replicate the results? A lot of people talk about faster than light travel and P=NP computations but it all sounds very sci-fi and not very serious. Could neutrinos be sent and measured accurately and reliably enough to allow for communication somehow, or somehow otherwise make dependable use of their (potential) superluminal features, or is this of purely theoretical value? Not to diminish the significance of it, but it’s a part of the debate that I’ve only seen uneducated hyperbole on

    Also, many thanks for a very informative and accessible blog. Your work is much appreciated 🙂

    1. Thanks for the positive feedback.

      I really think it is too early to answer your question. The answer will depend on how and whether we have to modify our understanding of space-time and of causality. And it will also depend on technological limitations, some of which it may be too early to imagine, but others of which are obvious.

      If OPERA is correct, and the supernova 1987 measurements are right (as seems certain), any superluminal neutrinos must carry very high energy (each neutrino carrying motion-energy 20 — 40 times larger than the mass-energy of a proton) and cannot be made without a big particle accelerator. Then you must make billions of billions of them to observe even one in a huge detector at a distance of 730 kilometers. This makes them practically useless for communication, either across the earth or across space, and probably useless for any other technological application in the near or medium future..

      Neutrinos may be very useful, however, as a tool in scientific experiments and related applications. That’s true even if OPERA’s result is wrong. They may be useful in geology, since they can cross the earth. They will eventually be useful in understanding how Type-II (core-collapse) supernovas work, as they emerge in abundance right after the core-collapse takes place. And high-energy neutrino astronomy is just beginning — neutrinos from cosmic rays might give us a new window into aspects of the universe. Where those investigations might lead, and how neutrinos’ unexpected properties might play in to those investigations, no one knows.

      History does not suggest that humans are good at predicting the technological future that arises from scientific discoveries. I don’t think your cell phone and its quantum mechanical transistors could have been predicted when it was first realized, just over a century ago, that an electron could orbit an atom without radiating away its energy (in contradiction to pre-quantum electromagnetism.)

  24. How likely is it that the effect might be due to us just not understanding the beam characteristics. For instance, the majority of neutrinos emitted at the leading edge vs. the trailing edge of the beam might be at different angles which would skew the fit the OPERA guys make for the incoming neutrino beam.

    1. I think this is potentially a serious issue, but am not knowledgeable enough to comment wisely. You really have to be a proton- and neutrino-beam expert to understand all the subtleties.

        1. I certainly don’t claim to be the only person to think of any of the possibilities I’ve mentioned on this site. I have no particular special knowledge of this topic that would make it likely that I would think of something that my many very smart colleagues would not think of too. But they too are likely to be wrong in most of their guesses.

  25. About OPERA – combining systematic errors in quadrature – I think just maybe some of the electronics timing errors may be correlated (think of an extreme case where there is a solar storm that causes EM interference all over the place), and so that should be looked at closely.

  26. Maybe I will read about my still open question attached to a previous post (concerning QG and superluminous neutrinos) in the remarks on theoretical conclusions in the last part later ;-P …?

    I like very much how things are nicely organised and represented here 🙂

  27. Hi Matt:
    Are neutrinos with any mass consistent with the SM? If neutrinos have mass do they can mass via the Higgs field? If not why not? Thanks for any response.

  28. Hi Matt
    I was thinking the following: the sky is so wide with so many stars. How come they knew that these tiny particles from this supernova explosion came from exactly the star that they were looking into. Because it may be that these neutrinos came from another star which could be a couple of light years behind the supernova they were analyzing? How did they differentiate between the supernova and the rest of stars?

    1. Most of the time, stars emit only very low-energy neutrinos, at a relatively low and very steady rate.

      But when a Type-II supernova explodes, it emits far more neutrinos in ten seconds than the entire galaxy that contains it! And these neutrinos are far more energetic than the ones that are steadily emitted from a star, and are therefore more easily observed.

      So if your eyes could see neutrinos, the supernova would be blinding. It would be a flash brighter than all the other stars in the sky combined. In other words, you can’t miss it.

      And here’s the evidence: look at the second figure in http://ircamera.as.arizona.edu/NatSci102/NatSci102/lectures/supernovaremnants.htm . Each black dot is one neutrino detection; the horizontal axis is time (over two minutes) and the vertical axis is the energy of each detected neutrino. You see a steady drizzle of neutrinos with about 5 MeV of energy (an electron has mass-energy of 1/2 an MeV, a proton has a mass-energy of 1 GeV = 1000 MeV). Some of those are coming from all the stars across the universe, many of them from the sun; others come from cosmic rays. Then suddenly there is a flurry of neutrinos, carrying much higher energies than usual… and then, back to the drizzle. That flurry was unprecedented, and nothing like it has ever occurred since. For that 10 seconds, whatever object emitted those neutrinos outshone all the stars in the universe. Three hours later — roughly the expected delay due to the fact that the star has a size and the shock waves need time to get to the edge before we can see the star glow — the supernova first became visible in camera images.

      So we have an unprecedented flurry of neutrinos, and 3 hours later we see the beginning of the brightest supernova of our lifetimes. And our understanding of supernovas predicts a several-second blast of neutrinos, with energies of tens of MeV, to precede a glowing supernova by a few hours. So… the case, though circumstantial, seems very, very tight.

  29. Hello Matt,
    I have been reading a lot of articles trying to better understand this phenomenon. I seem to remember somewhere that only the tau-neutrino measured faster than light results, is that correct? If so, what would that be indicative of? Also, can you please explain how a neutrino can go the speed of light given that it has mass?

    Thank you for your time.

    1. actually the beam is mostly “muon-neutrinos” (neutrinos that were created along with an anti-muon in the decay of a pion.) A few will have changed into tau neutrinos while traveling through the earth. I think to discuss the significance requires a longer commentary than I can give here, because it requires discussing some subtleties about neutrino identity.

      A particle that has mass can get as close as you want to the speed of light, in Einstein’s theory, though it never quite gets there no matter how much energy you add. Somebody earlier asked about acceleration of particles and you may want to read my response; look for the question from “aks” on this page: http://profmattstrassler.com/2011/09/20/supernovas-and-neutrinos/

      Of course if this experiment is right, Einstein’s theory requires modification, and so my answer to your question will also require modification.

  30. Hi, first of all thanks Matt for this amazing blog.
    Maybe it’s a silly question but, why do they have to set the origin and detector so far apart? I mean, there’s no way to do the same experiment in a lab with a detector let’s say 20 meters away from the neutrino beam origin?
    I’m sure there’s a good reason for that but I just can’t figure out the answer 🙂


    1. The experiment was actually designed to do something else… that’s why the distance is so large. The speed-of-light measurement was not the focus — the OPERA experimenters, like everyone else, fully expected that their measurement of neutrino speed would find they travel at the same speed as light, up to the precision of the experiment.

      Instead what the OPERA experiment was (and still is) mainly supposed to do is this: (you may want to read my posts linked above on neutrino beams and neutrino detection)

      *make neutrinos by making lots of pion particles, which decay to antimuons and neutrinos,

      *let those neutrinos travel a long distance through the ground,

      *measure a few of those neutrinos hitting an atomic nucleus in a detector

      *see if any of the neutrino-nucleus collisions turns the neutrino into a tau particle rather than a muon.

      The old expectation, back about 30 years ago, was that a neutrino that is initially produced along with an anti-muon could in a later neutrino-nucleus collision only turn into a muon. We’ve learned in the past 30 years that this isn’t true. Instead, depending on how far the neutrinos travel and whether they travel through some amount of matter (such as rock or the sun), their properties will change. One good way to detect and study this change, and in particular the possibility of making taus instead of muons in neutrino-nucleus collisions, turns out to be: make a beam such as the one CERN aims at OPERA, and have the neutrinos travel hundreds of kilometers before you measure them.

      The question of whether the neutrino-speed experiment can be done on a much shorter distance scale is a very interesting one and I assure you experimental physicists in many places are thinking hard about it. 20 meters is too short because the precision timing required would be beyond current methods. But a kilometer or so might be about right. I’m sure we’ll hear experimental proposals of this sort soon enough.

  31. Aside from the question of whether the results are true – once people accept that there were no outright experimental mistakes made, there seems to be a consensus that it would challenge current physics (theory) and adaptions would have to be made.
    I wonder about one thing in particular: Is changing/expanding on special relativity the only option (because the speed of light is what everybody is talking about) or are there other relevant (and established) theories that, if untrue/incorrect, could explain the results?
    As an example: maybe special relativity is correct and the speed of light wasnt exceeded, but we have a flawed understanding of neutrino physics that impacted the results… etc. (i am not sure if I make myself completely clear)). The c-limit seems to be so well established that I wonder if there are less-well tested theories relating to the experiment that – once scrutinized and expanded/reworked – could also explain the result.

    1. I am sympathetic to your question, but every time I try to answer it I find myself writing nonsense. So let me simply say: I suspect the answer to your question is no, but am not sufficiently clear-headed yet to want to commit to that answer.

      1. Prof. Matt, what is your view about extra dimensional theories ? They say neutrino can take extra dimensional shortcut and defeat light. Although what i saw was for sterile neutrinos — i.e. neutrinos of an exotic type that, unlike the three types of neutrinos in the Standard Model, are not affected by the weak nuclear force. [Question edited by host for clarity]

        1. I need to make sure my answer is right before I give a definitive one. As a preliminary remark, however, let me comment that the three types of neutrinos in the Standard Model of particle physics are inextricably linked with the three types of charged leptons: the electron, the muon and the tau. [You can see that in my article http://profmattstrassler.com/articles-and-posts/particle-physics-basics/the-known-particles-if-the-higgs-field-were-zero/ .] It would naively appear impossible to allow for Standard Model neutrinos to take such a shortcut without allowing charged leptons to do the same — and there are powerful experimental constraints on electrons and muons traveling faster than light. A charged particle traveling faster than light generates a “photonic boom” (Cenenkov radiation),a light-analogue of a sonic boom, and this causes the charged particle rapidly to lose its high energy. (We see this in water, which slows light down a bit; very-high-speed charged particles in water can travel faster than light in water, and they do emit a photonic boom.) But we observe very high-energy muons and electrons; they haven’t lost their energy. So it appears naively that for neutrinos to do what you suggest could only be the case for exotic (specifically “sterile”) neutrinos, whereas the neutrinos created in the CERN neutrino beam were of a standard type. That said, there are a few loopholes in this argument that I have not yet closed. Thus it would seem unlikely that extra dimensional theories provide a natural solution, but I am not yet at the point where I personally would swear to that. Other experts may know more.

      2. Thanks for your nice reply. I have just started learning neutrino oscillation now. Probably i will work on sterile neutrino oscillation on MS thesis. Did not read those papers yet. Still a naive. So just asked you 🙂

      3. Prof. Strassler– you say: “A charged particle traveling faster than light generates a “photonic boom” (Cenenkov radiation),a light-analogue of a sonic boom, and this causes the charged particle rapidly to lose its high energy.”

        Can you clarify this?

        One– when I look on wikipedia, it defines Cenenkov radiation as “electromagnetic radiation emitted when a charged particle (such as an electron) passes through a dielectric medium at a speed greater than the phase velocity of light in that medium”. But you seem to be describing Cenenkov radiation emitted by charged particles outrunning light via “shortcut-taking” (ie whatever the neutrinos at OPERA are doing) (or possibly not doing), rather than outrunning via light being slowed by a medium. You say “we observe very high-energy muons and electrons; they haven’t lost their energy”. If these high-energy muons and electrons you refer to have been traveling through space, or were created inside of an accelerator, then they have not been moving through a dielectric medium, or any medium at all; they have been moving through vacuum. Is there some difference between the Cenekov radiation you’re talking about here, and the Cenekov radiation wikipedia says occurs in “a dielectric medium”? Or does your argument refer very specifically to muons and electrons that are known to have passed through earth’s atmosphere (Google says air is dielectic?!) before hitting the detector, yet still hold energy unrealistic if they were Cenenkov-radiating while moving through said atmosphere?

        Two– do I assume that neutrinos moving faster than light via “shortcut-taking” would *not* emit this Cenenkov radiation– because they are not charged?

  32. Maybe this is too candid even for this blog but, has anyone tried to see if there had been a burst months or even years before 1987A supernova ?

    1. I’m not absolutely sure — we’d have to ask the people who ran those experiments — but I think the answer is probably yes. After the 1987 supernova, people started combing through neutrino data on a regular basis looking for bursts. There is always the possibility that a supernova will occur in a part of the galaxy shrouded by dust, in which case it might *only* be seen through its neutrinos. So I’m pretty sure people would also have gone back in their data as far as possible. Of course there is always the possibility that neutrinos came in when one or more of the detectors was under repair. So maybe there would be a loophole. And of course the neutrinos might in principle have straggled in over days and caused no burst at all.

      However, for reasons that I will explain later, it appears that there is a very strong experimental case that neutrinos of different types do not travel at significantly different speeds. [I have to make sure that statement has no loopholes before I expound on it.] If that is true, then I don’t think you could have one class of neutrinos arriving early and another class arriving on time.

  33. We know that since 1988 the Large Magellanic Cloud were blow up and we also received neutrinos and fotons just in same time. If this opera-theory be true, do not we need detected neutrinos in 3 years before, since it is 160000 light-years distance from Earth?

    1. You argument is sensible (and see my post on Supernovas and Neutrinos linked above which discusses this in detail.) But many effects in physics depend on a particle’s energy. Since the energies of the OPERA neutrinos were about 500 times larger than those of the neutrinos from the supernova, it is possible that the OPERA measurement and the supernova measurement actually are consistent, as long as the Einstein-violating effect grows rapidly with energy. This of course implies that physicists should repeat the OPERA measurement at both higher and lower energies, to observe whether the effect grows with energy.

  34. Hi Matt,
    Following the OPERA’s presentation at CERN last week, one of the audience at made a point that I found quite interesting. Namely, he was pointing out that the positions of the underground labs – relative to a point above them on the surface (needed for an accurate baseline measurement) – had only been made twice and that both used the same technique. He finished his point with a phrase that made my laugh [paraphrasing]: “given the implications of your results, why don’t you dig a bloody hole?!” I presume that this was suggested to allow a laser (or something similar) to be used to more accurately measure the lab(s) position(s), relative to the surface. Do you think that a mismeasurement of the baseline like this could account for the apparent early arrival of the neutrinos and how convinced are you of the method that they used to make this measurment?

    1. I do think this is something to be concerned about, among many other subtle points, but I am in no position to have an opinion. As I recall, they used external surveyors to help them out, so there is dependence on their quality. As for digging a hole — again, you’d have to know about the geometry and geology of the site, but I assume that would be prohibitively expensive. It might be cheaper to build a simpler and dedicated experiment that simply avoids this problem and many others.

Leave a Reply


Buy The Book

A decay of a Higgs boson, as reconstructed by the CMS experiment at the LHC


A quick note today, as I am flying to Los Angeles in preparation for and other events next week. I hope many of you were

POSTED BY Matt Strassler

POSTED BY Matt Strassler

ON 04/09/2024

POSTED BY Matt Strassler

POSTED BY Matt Strassler

ON 03/05/2024