Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Do Tachyons Exist?

  1. Aug 16, 2015 #1
    For this question, I should probably clarify my question. I want to know, more or less, what the general consensus is on the existence of tachyons. What do most scientists in fields that are relevant to this question think about tachyons, and - as a side question - what might their reasoning be for their positions?
     
  2. jcsd
  3. Aug 16, 2015 #2
    Tachyons are hypothetical particles which can only travel faster than light.
    They are not part of the standard model of particle physics and there is no experimental evidence of them.
    If they do exist then the whole ediface of SR and GR would be wrong, yet there are mountains of persuasive evidence confirming the predictions of relativity.
    As far as I know they mostly are considered as a mathematical artifact.
     
  4. Aug 16, 2015 #3

    mfb

    User Avatar
    2017 Award

    Staff: Mentor

    In particle physics, they are considered to be even more exotic than those searches.
    I don't think anyone really expects them to exist.
     
  5. Aug 17, 2015 #4

    Demystifier

    User Avatar
    Science Advisor

    Tachyonic fields are not really so exotic as usually thought. For example, the Higgs field in the standard model before symmetry breaking is a tachyonic field, in the sense that the mass term in the Lagrangian has a negative sign. But such a tachyonic field configuration is unstable, so it quickly roles down to the minimum of the potential, thus breaking the symmetry and becoming a stable a massive Higgs.
     
  6. Aug 17, 2015 #5

    Avodyne

    User Avatar
    Science Advisor

    Yes, but not in the sense that it goes faster than light, which is what people usually mean by a tachyon.
     
  7. Aug 18, 2015 #6

    Demystifier

    User Avatar
    Science Advisor

    If you solve the classical field equation in the vicinity of the maximum of the Higgs potential, you will see that the field does propagate faster than light.

    More precisely, the oscillatory modes propagate faster than light. There are also modes exponentialy growing with time, which do not propagate faster than light. Since the exponential modes describe instability which results in settling down to the minimum of the potential, it is believed that this exponential not-faster-than-light solution is the physical solution. Yet, it is not clear if there is any physical principle which could generally prevent a realization of the oscillatory faster-than-light solutions.
     
    Last edited: Aug 18, 2015
  8. Aug 18, 2015 #7

    Avodyne

    User Avatar
    Science Advisor

    Do you have a reference for this claim?

    I would expect that the classical initial-value problem does not allow signals to propagate faster than light, even if the initial data includes modes with wavenumber ##k>|m|## (which is what I assume you mean by "oscillatory modes").

    EDIT: here is paper that states that there is no superluminal propagation for the classical initial-value problme even with negative mass-squared (see p.8):
    "No superluminal propagation for classical relativistic and relativistic quantum fields"
    John Earman
    http://www.sciencedirect.com/science/article/pii/S1355219814000811
    pdf: http://philsci-archive.pitt.edu/10945/1/NSP_SHPMP_Final_Version.pdf
     
    Last edited: Aug 18, 2015
  9. Aug 19, 2015 #8

    Demystifier

    User Avatar
    Science Advisor

    Most of what I claim can be easily derived from the tachyon dispersion relation and assumption that group velocity is the velocity of propagation.

    Yes, this is what I mean by oscillatory modes. For such modes the group velocity is larger than c. So why do you expect that classical initial-value problem does not allow signals to propagate faster than light?
     
  10. Aug 19, 2015 #9

    ShayanJ

    User Avatar
    Gold Member

    What about this?
    It says:
     
  11. Aug 19, 2015 #10

    Demystifier

    User Avatar
    Science Advisor

    I agree with this. Indeed, it is easy to understand it intuitively. For tachyons, space and time exchange their roles (this is the most obvious in 1+1 dimensions). Normal massive matter is local in space but nonlocal in time (matter lives "forever" due to conservation in time), and has subluminal velocity dx/dt<1. Likewise, tacyon matter is local in time but nonlocal in space, and has subluminal inverted velocity dt/dx<1. The Cauchy surface must be spacelike for normal matter ("initial" condition), but timelike for tachyons ("boundary" condition). Normal conscious beings experience "the flow of time", while tachyon conscious beings would probably experience "the flow of space".

    The real problem appears when you have an interaction between normal and tachyon matter. Should the Cauchy surface be spacelike or timelike? Locally it doesn't matter (Cauchy-Kovalevska theorem provides the local existence of a solution near the hypersurface), but globally the Cauchy problem is not well defined. This means that the initial/boundary condition cannot be arbitrary, because otherwise you can obtain inconsistencies (e.g. the grandfather paradox).

    The above were features of classical fields, but quantization of tachyons leads to additional challenges. Canonical quantization gives a special status to the time coordinate, but for tachyons time and space should exchange their roles. Even if this is not a problem in 1+1 dimensions (at least if there are no interactions with normal matter), this is very problematic in 3+1 dimensions because, even in a fixed Lorentz frame, now we have 3 natural "canonical momenta" and 3 natural "Hamiltonians". How to perform canonical quantization with this? An option is to do covariant path-integral quantization, but it only works for calculation of n-point functions. If you want to calculate physical probabilities via something like LSZ formula, the problems of canonical quantization are turned back. For instance, should probabilities be conserved (in the sense of unitarity) in time or in space? Both options are problematic for tachyons.

    Finally, let me emphasize that all this is not purely academic. Suppose that I want to describe the Higgs field before symmetry breaking, i.e. at sufficiently large energies. In that regime the Higgs field looks tachyonic, it is a quantum field, and it interacts with non-tachyonic matter. If one wants to describe the actual experiment with an accelerator such as LHC, the quantization and all the theory can still be done in the laboratory frame, which may be sufficient for a comparison with the experiment. But the theory (describing quantum interacting tachyons in the laboratory frame) will not be Lorentz invariant. Perhaps this is not inconsistent in an instrumental interpretation of quantum theory (according to which the role of quantum theory is not to describe nature as such, but to describe what can we say about nature in a given experimental setup), but it is at least disturbing.
     
    Last edited: Aug 19, 2015
  12. Aug 19, 2015 #11

    Demystifier

    User Avatar
    Science Advisor

    I've just discussed with a colleague the idea of producing such tachyons in LHC. We finally agreed that such a tachyon could be described as a complicated bound state of a large number of massive Higgs particles, analogous to a QCD glueball made of a large number of gluons. To stress the analogy, he invented a suggestive name for such a tachyon - the higgsball.

    Another related thought. Even if such tachyons cannot in practice be created in LHC, perhaps something similar could be created in condensed matter. What I have in mind is an acoustic tachyon, an excitation moving faster than sound described by a tachyon dispersion relation associated with the velocity of sound. The dispersion relation is Lorentz invariant with respect to Lorentz transformations with an invariant velocity of sound. The violation of Lorentz invariance due to quantization would have a clear physical interpretation; there is a preferred frame, the one with respect to which the condensed-matter material is at rest.
     
    Last edited: Aug 19, 2015
  13. Aug 19, 2015 #12

    Avodyne

    User Avatar
    Science Advisor

    This disagrees with textbook QFT. A theory with a Lorentz-invariant lagrangian density that results in spontaneous symmetry breaking (such as the Standard Model) remains Lorentz invariant (at all energies and for all processes), and there are no tachyons (that is, localized excitations of any sort that move faster than light).
     
  14. Aug 19, 2015 #13

    mfb

    User Avatar
    2017 Award

    Staff: Mentor

    A preferred frame in solid matter is not really surprising.

    A technical remark: a group velocity larger than c is possible in special relativity if there is nonlinear absorption or emission. This has been observed with light, see the references there.
     
  15. Aug 20, 2015 #14

    Demystifier

    User Avatar
    Science Advisor

    That is a valid objection, but see post #11. A tachyon excitation should be a non-perturbative effect that cannot easily be described by the standard textbook techniques, very much like a glueball in QCD. And I am not saying that it is easy to create such a tachyon in LHC collisions. I am only suggesting that it should be possible in principle, even if the probability of creation is too small to consider this possibility more seriously. Certainly the possibility of such a tachyon is, at the moment, too exotic to be discussed in textbooks. But it can be an interesting idea for a more detailed research paper. And it sholud be appropriate to discuss such exotic stuff at the Beyond the Standard Model forum.
     
    Last edited: Aug 20, 2015
  16. Aug 20, 2015 #15

    Demystifier

    User Avatar
    Science Advisor

    Of course it's not. But what if QFT in condensed matter is how QFT should generally be interpreted? What if all SM particles are only quasi-particles like phonons? What if Lorentz-invariant QFT of the standard model is only an effective theory, while the fundamental theory has a preferred frame? That could solve many conceptual problems like those concerning tachyons, non-local hidden variables for EPR correlations, the problem of time in quantum gravity, UV completeness of quantum gravity (Horava theory), and so on ...

    Interesting!
     
    Last edited: Aug 20, 2015
  17. Aug 21, 2015 #16
    We have enough trouble trying to detect conventional particles like neutrinos. If tachyons existed, we and everything around us might have them flying through us, like neutrinos do, and never know it.

    Everything conventional has a speed of light speed limit. Tachyons don't so may not be particle, wave or anything else we know about. I don't think anything we currently know about could detect them, unless maybe one day a neutrino detector detected a collision which went way off the normal energy scale.
     
  18. Aug 22, 2015 #17

    phinds

    User Avatar
    Gold Member

    Actually, tachyons DO have a "speed of light speed limit", it's just that it's a minumum speed limit rather than a maximum which is what normal matter has.
     
  19. Aug 22, 2015 #18
    I know that an idea is that the more energy they have, the slower they go, till they reach light speed but that just does not make sense. If tachyons have any conventional energy, they will be limited to conventional speeds. I would find it more believable if they, like photons, had a set speed.
     
  20. Aug 22, 2015 #19

    phinds

    User Avatar
    Gold Member

    And what do you propose that "set" speed is? It can't be c because they have to travel faster than c. Should it be 1.1c? How about 9.62c? How would you justify such as set speed?
     
  21. Aug 22, 2015 #20

    mfb

    User Avatar
    2017 Award

    Staff: Mentor

    Such a fixed speed is impossible: relative to what? Spacetime transformations allow the existence of exactly one fundamental speed that appears the same to all observers. We found this speed, it is typically called "speed of light" as light in vacuum moves at this speed.

    There are many observed things that don't appear to make sense when you first hear about them. Everyday experience is a bad judge.
     
  22. Aug 25, 2015 #21
    If you're going for what would make it make sense, I would assume that the "imaginary energy," however that would manifest itself in reality, would be more or less equivalent to negative "conventional" energy, meaning that by adding conventional energy, it would lose imaginary energy and thus get slower, and by adding conventional energy, it would lose imaginary energy. Unfortunately though, considering how much normal science can contradict common sense, if tachyons do exist, they probably make no sense without a lot of time and effort having gone into understanding them, so I expect that this isn't actually the case.
     
  23. Aug 25, 2015 #22
    Your comments have been very interesting to read, and I sincerely appreciate your help, but I did find much of this very difficult to understand, and I feel as though I only barely understood most of what you said, so would it be possible for you to explain this a slight bit more simply? While I probably have a better understanding that many in my position due to my strong interest and enjoyment of science as a hobby and future carrier path, I still only have as much formal education as a High School Senior (and only a week of school as a senior so far). Still, I understand if this is not a topic that can simplify very much or very easily.
     
  24. Aug 26, 2015 #23

    Demystifier

    User Avatar
    Science Advisor

    In the text you quoted I mentioned a dozen of otherwise not directly related concepts. To explain those to someone with high school education, each would require a few teextbook pages of thorough explanation. I hope you understand that I cannot provide such explanations here. But I can answer a more specific question, if you have one.

    For a start, you can read more about the Lorentz interpretation of relativity, e.g. here
    https://www.physicsforums.com/threads/lorentzian-relativity.298185/
    http://www.worldsci.org/pdf/abstracts/abstracts_6205.pdf [Broken]
     
    Last edited by a moderator: May 7, 2017
  25. Aug 26, 2015 #24

    Urs Schreiber

    User Avatar
    Science Advisor
    Gold Member

    I agree with what Demystifier says, this here is just to amplify that the issue is one of which vacuum one is perturbing about. Perturbation theory, by construction, is (up to issues of resummation) the Taylor expansion of what really ought to be a globally defined function (the S-matrix) around some point of its parameter space. Usually one takes that point to be the minimum of the relevant potential. Then all particle quanta, which are the small oscillations around that minimum, have positive mass squared, determined by the second derivative of the potential at that point: at a minimum, the first derivative vanishes, the second is positive and gives positive mass squares to particles.

    Now, even though it is often not useful, it is nevertheless mathematically possible to instead look at the peturbation series around not a minimum but a maximum of a potential. Then its second derivative is negative there, and accordingly the mass squared of the small fluctuations around that points come out negative. These are the tachyons. Their presence signifies that the perturbation is not done around a potential minimum, but around a maximum. This is where the example of the Higgs particle comes in. Its Mexican hat potential does have a maximum, and even though one usually does not do perturbation theory around this maximum, because by design that's not the point of interest, it is in principle possible and doing so will show that there is a tachyon in the spectrum at that point, signifying the tendency of the theory to roll down away from that point.

    Notice that the theory doesn't change as we change the point where we base its perturbation series, just the approximation to it changes.

    One of the best studied cases of this is the open string tachyon. Consider it as an academic example if you wish, just to help understand the situation. In open string theory there is a tachyonic excitation, and it was early on voiced by Ashoke Sen that this hence means that the open string perturbation series is the perturbation about a potential maximum (namely that of the space-filling D25 brane, for what it's worth), hence that the stable/true vacuum of open bosonic string theory should be elsewhere. Eventually that picture was verified by detailed computations in open string field theory, one of the big computational successes in the field. References are here.
     
    Last edited: Aug 26, 2015
  26. Aug 26, 2015 #25

    Avodyne

    User Avatar
    Science Advisor

    You can call these excitations "tachyons" if you like, but they do not move faster than light. The same is true of the open-string "tachyon".
     
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook