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

What are unified theories trying to match at lower energies?

  1. Jan 15, 2010 #1
    What do the phenomenologists of unified theories (strings, loops etc) try to match at lower energies when there are so many conflicting theories lying around at these energies?
    I understand that certain things should be there in any model - like chiral matter, three generations etc - but a lot of things still remain unsettled (like right-handed neutrinos, the GUT gauge group etc).
     
  2. jcsd
  3. Jan 15, 2010 #2

    tom.stoer

    User Avatar
    Science Advisor

    I know about two approaches regarding unification, string/M theory and non-commutative gravity (a la Connes). LQG is certainly not an approach to unification but to quantum gravity "only".

    Regarding strings: I guess that F-theory model building tries exactly what you have in mind, namely to tailor low-energy effective and phenomenologically viable theories. You should have a look at http://arxiv.org/find; try author = "Vafa", title = "F-theory".
     
  4. Jan 16, 2010 #3
    I have looked at papers on string model building but I get confused as to the phenomenology that is matched at low energies, since I see a lot of conflicting (partially, if not fully at loggerheads) models at low energies.

    So, to rephrase my question:
    What are the 4-dim low energy phenomenological features (other than the experimentally verified ones of the SM) which any compactified model should satisfy at the bare minimum, in addition to making new predictions?
     
  5. Jan 16, 2010 #4

    tom.stoer

    User Avatar
    Science Advisor

    I do not understand. As there are (to my knowledge) no experimental results beyond the standard model, I do not see what we can expect from unified theories.

    There are three important points:
    1) post-dict well-known SM results
    2) predict new physics beyond the SM
    3) predict assumptions / axioms of the SM

    Regarding 1) alternative theories still have problems to derive all known results / not to run into conflicts.
    Regarding 2) there are a "predictions" like SUSY, unfortunately rather unspecific (at which energy scale? how to break it? ...)
    Regardig 3) there are a lot of assumptions in the SM that I would like to see as a result of a new theory: 3 spatial dimensions, 3 fermion generations, U(1)*SU(2)*SU(3) gauge group of the SM, gravity, Higgs, ... to mention a few.

    My expectation is that sooner or later we will be able to understand at least some of these inputs of the SM as the results of a more fundamental theory; but currently I do not see any candidate at the horizon (personally I don't believe that it's string theory).

    Tom
     
  6. Jan 16, 2010 #5
    Thanks Tom. Few more questions on the three points...

    1: The number of independent well-known (aka experimentally verified) results of SM depend on the model used in GUTs etc. And the higgs mechanisms and the scales at which they occur have also not been decided conclusively (like SO(10) might break via the Pati-Salam model or not and so on). Even the particles predicted in different GUTs differ (like right handed neutrinos in SO(10), differing lifetimes of proton decay and so on). In such a situation, what can the unified theory's phenomenologists target at low energies?

    2: If there had been concrete features that the unified theories had to have at the very least at low energies, we could then say that the rest of the spectrum are new predictions. But lacking concrete features (except maybe a general SUSY-GUT scale of 1016GeV, three families of quarks and leptons, the GUT scale higgs/yukawa etc), the rest of the spectrum of the compactified theory will not be taken seriously as new predictions. As for SUSY breaking scale and mechanism, it is again something that SM phenomenologists should worry about right?

    3: Three spatial dimensions are not hard to get and so are the three generations and gravity (at least SUGRA).
    The U(1) x SU(2) x SU(3) gauge group of the SM is probably not what compactified models should aim for - I would expect them to aim for the SUSY-GUT gauge group (which, at the moment, is undecided).
    Higgs/Yukawa - at least the one that is responsible for the GUT break at 1016GeV - is something that can indeed be a potential target for compactified theories.

    And LQG is not aiming for unification? I thought they were vying for the same pie as string theorists! So, is their idea something like quantize gravity for now and later try for a gauge group that would include the SM and the quantized gravity?
     
  7. Jan 17, 2010 #6

    tom.stoer

    User Avatar
    Science Advisor

    If a "candidate" theory makes predictions different from the SM and disproved by experiment you can either modify it - or throw it away
    Why do you thing that it's easy the get three dimensions? I know of no theory which predicts three dimensions. If you theory produces a SUSY-GUT, there's still the problem tobreak it down to SM in a realistic way. Eventually this has to be understood as well; deriving SUSY does not help in a world w/o SUSY.

    LQG is not aiming for unification. It is compatible with all known gauge symmetries and perhaps even with SUGRA. It provides something like "unification" as it implements gravity as a gauge theory, so at least the framework is harmonized.
     
  8. Jan 17, 2010 #7
    1- if LHC does not find evidence of SUSY, nor other experiments, is there a physics reason to continue researching it?
    2- if the braiding scheme via Bilson Thompson does work using framed graphs, would you regard this as LQG unification?
     
  9. Jan 17, 2010 #8

    tom.stoer

    User Avatar
    Science Advisor

    Regarding 1) I would say both yes and no. Yes because maybe SUSY exists (unbroken) at higher energies. No because it makes no sense to search for SUSY at 10, 20, 30, ... TeV.
    Regarding 2) I would say it's wishful thinking - but YES, that would be the deepest unification I can think about. It's not that you use higher and higher symmetry groups, dimensions, SUSY etc.You just quantize ordinary space-time, braid it - and all particles and forces emerge from it! Are there any new results in the last month?
     
  10. Jan 20, 2010 #9
    @ tom.stoer

    Pardon me, I did not mean predicting 3 dim or 3 families, but meant engineering them from a predicted higher dim. I think the requirement for a unified theory to predict the axioms of SM can be put on hold for now - will first look for 'how' and then think about 'why'.

    The problem of how SUSY-GUT reaches SM is far down from the realm of unification energies . So, I would expect them to be the subject of research of 'physics beyond SM' than 'unified physics' - in my earlier example, SO(10) breaking via Pati-Salam model or not is a question for SM phenomenologists than string phenomenologists.
    In addition, as I have been saying, the SUSY-GUT is also undecided and there are a lot of proposals there. So, what would the model-builders of a candidate theory of unification aim for at the lower energies? In other words, till the time the 'why's are answered, what benchmark could be used in evaluating low energy models engineered from unification theories?
     
  11. Jan 20, 2010 #10

    tom.stoer

    User Avatar
    Science Advisor

    I am not so sure what you mean by engineering. Do you have something in mind like "take a stack of xyz branes + blablabla" and check something like "SM + abc" comes out AND "abc can be hidden somehow"? If this is what you mean then you are "only" looking for concepts how to engineer SM from something else.

    First I would expect that the SM comes out (unfortunately this is not the case in F-theory so far,even if you are quite close to it)
    Second I would like to understand how to hide abc; should be rather natural - whatever that means.
    Third I would like to see some hidden / higher-order effects from xyz or abc. This could be something like the restriction regarding the Higgs mass coming from higher loops (at tree level the Higgs mass is completely arbitray, but at higer loops there are constraints which are subject to experiment at the LHC). Or you will find something like proton decay which killed SU(5) and SO(10) GUT. Or you will find long range forces (which basically kills many string-inspired models as they quite often predict SM * U(1) * U(1) * ... or SM plus massless scalars.

    I don't see any generic prediction or something like that. It depends on the model you start from (e.g. xzy) - and it depends on abc, of course.
     
    Last edited: Jan 20, 2010
  12. Jan 20, 2010 #11

    arivero

    User Avatar
    Gold Member

    Last edited by a moderator: Apr 24, 2017
  13. Jan 20, 2010 #12

    arivero

    User Avatar
    Gold Member

    Hmm??? I am not even sure if they are easy to engineer, nor to predict. Al least, in accepted peer reviewed theories :biggrin:
     
  14. Jan 21, 2010 #13
    @ tom.stoer

    ya, thats what I had in mind when I said engineering low energy models. I agree with your second point on hiding the exotic particles naturally and the third point on fixing the higgs mass, but not fully (or perhaps I have wrong ideas about what is expected) with first point:

    Is it really expected of string theorists to be the first to come up with an undeniable GUT model and also a mechanism for breaking it down? I was expecting the GUT-scale and astroparticle phenomenologists to be the first to fully agree on a few (~ 2-3?) potential GUTs and their break down mechanisms (rather than the open endedness that I think I am seeing - even proton decay is not fully ruled out yet). Then string phenomenologists could aim for these few potential GUT models along with predictions in each case about GUT higgs mass, candidate dark matter/energy etc which could be the deciding factor in fixing one single GUT theory.
     
    Last edited: Jan 21, 2010
  15. Jan 21, 2010 #14
    @ arivero

    In intersecting brane models, for example, it boils down to engineering the two fractional 3-cycles in the 6 compact dimensions to intersect 3 times. Its surely not a prediction (yet or maybe never). I have seen 3 generations in other compactifications too, but I dont remember how they were engineered. And, sure, none of them had no unwanted/missing particles in them - if not, such a result would not have gone unnoticed.
     
    Last edited: Jan 21, 2010
  16. Feb 1, 2010 #15
    From my limited exposure, I get the feeling that the theories at low energies (lower than string scale) are developed only to the extent of using them to classify 4D models constructed down from string theory as those that are obviously wrong and those that could potentially be right. I don't see this helping the "string or no-string?" discussion at all.
     
  17. Feb 1, 2010 #16

    tom.stoer

    User Avatar
    Science Advisor

    I am not an expert but I am expecting the following:
    - try to get as closed to the SM as possible i.e. reduce the unwanted features (like in F-theory)
    - understand what the essential features are to reproduce the standard model
    - create similar models closed to the standard model and understand this local region of the "landscape"
    - ...

    I guess entering the second or third stage would be called the third superstring revolution; working on the first stage may by boring. I really cannot say whether it's worth to continue or better to give up.
     
  18. Feb 2, 2010 #17
    @ tom.stoer

    I still do not understand why SM is the target.
    Let me put it this way: Even among GUTs, all the candidate theories have SM somewhere within them - it is just how they break themselves to reach that SM (eg. SO(10) reaching SM through Pati-Salam model or not? What energy scales at which these breakings occur? etc) and other new extra predictions (proton decay, neutrino masses etc) that differentiate them.

    Similarly, for string theory, SM is contained in a ton of models - but, it is the new extra predictions that differentiates them and will even select The One among them. But, the problem is, there is still a debate as to the nature of these extra features (at lower ~ GUT scales). Till this is settled, there is really no one fixed criteria against which the string models can be matched and judged. So, the question of string theory being successful or not does not make sense.

    So, instead of debating this, shouldn't the topic of debate be how much we should fund research at Planck scales when there is so much pending work at GUT scales?
     
  19. Feb 2, 2010 #18

    Fra

    User Avatar

    IMO, I think this strategy risks missing some of the makeup of nature, and natures workings.

    If some of the "how", is developed on the fly like evolutionary self-organizing progression, the why is almost more important. In the extreme reduction, the how could simply be an abstract random walk where the more important thing could be to understand the feedback and selection mechanism of the self-organising random walker.

    The how is then a result of evolution, while the WHY amounts to understanding the mechanics of variation and selection.

    I expect the how, to follow from first principles at least to a reasonable extent. At least in the sense of probably evolutionary paths, in analogy (an even an extension/generalisation of) to how we understand life on earth.

    I want for something along the lines of what are the self-organising principles that could be responsible for the evolution of diversity we seen, given a starting point where differentiation of forces and 4 spacetime dimensions was no longer distinguishable to an inside observer? Could these principle explain what we see as an "expected outcome" out of such a self-organisation given enough time?

    For string theory to come near this, also the origin of the strings and their background needs explanation. They why strings, has to be answered. I won't accept less.

    /Fredrik
     
  20. Feb 2, 2010 #19
    @ Fra
    Sure that would happen.
    But, we have to find out how this universe exists before answering why it exists - anything else would be philosophy, not physics. ;)
     
  21. Feb 2, 2010 #20

    Fra

    User Avatar

    Maybe I can clarify a bit.

    Yes I think the "unification quests" by definition overlaps with philosophy in the first place :) Even thouhg physicistst may deny it. Why do we seek "unification" in the first place? Why is a coherent understanding better than a fragmented "collection" of a diversity of actual experimental results?

    I hold the position that how and why are not really that different. If you take a more information theoretic approach to things (like I do), the why, as in "confidence in opinon or expectation" clearly conceptually realted to "how" information is processed and stored.

    So when I ask "why" I do not mean it in the traditional philosophical sense, I rather mean to ask for a information measure of the confidence in a given expectation. So, why is spacetime 4D, translates to - how is information about the environment, processed and retained so as to yield a rational expectation of a 4D spacetime?

    For me why and how are related in this sense. Then again, string theory is not an information theoretic framwork, it's more old style building on a sort of realist picture. I find this unmodern and not satisfactory.

    /Fredrik
     
  22. Feb 2, 2010 #21

    arivero

    User Avatar
    Gold Member

     
  23. Feb 2, 2010 #22

    Fra

    User Avatar

     
  24. Feb 2, 2010 #23

    tom.stoer

    User Avatar
    Science Advisor

    I don't agree. I have not seen one single model (GUT, SUSY, string) which is not either completely unnatural (*) or simple wrong (**).

    Let's start with (**). If a model predicts something similar like the SM (where similar means that the predictions do not match the experimental results) then this model is wrong and must be thrown away.

    (*) is more difficult. If a model makes predictions A, B, C, ... which you really do like, and it makes predictions X, Y, Z, ... which are simple wrong in terms of (**) - but you can save the model by assuming this, hiding that ..., then I would call the model unnatural.

    If a model allowes for U(1)*SU(2)*SU(3) but w/o the correct chiral structure it is WRONG. If a model predicts this goup but in additions has unwanted U(1) factors it is WRONG. If you are able to hide these extra structures at higher energies etc., then you may be on the safe side. But if you look at attempts search for physics beyond the standard model the situation is really a mess: either the models are wrong in the sense of (**) (as proton lifetime is too short ...) or they are unnatural in the sense of (*) (because they essentially double physical objects like particles in a hidden sector).

    Neither do I agree with your claim that "all the candidate theories have SM somewhere within them", nor do I agree with "SM is contained in a ton of models". It always "something similar to the standard model" or "closed to the standard model" or that "could be the standard model".

    Because there is no physical evidence for one single experimental result beyond the standard model, the main focus must be reproducing the standard model; after this has been achieved (in a certain class of models) one can search for new physics beyond the standard model! Most candidate theories fail (failed) because they do (did) not reproduce the standard model . It is a waste of time to look for new physics beyond the standard model if you already know that the model in question is in disagreement with experiment.

    I do not say that doing it the other way round is wrong, I say that its ineffective as you have lost your guiding principles
     
  25. Feb 2, 2010 #24

    arivero

    User Avatar
    Gold Member

    Recently I have become worried about this point. Are we really sure about which chiral structure, if any, must we match?

    In the low energy, the standard model is not chiral, it is SU(3)xU(1). The chiral part is Fermi interaction, a non renormalizable object. And the Z0 boson has no role here, even.

    At high energy, is is true that we have U(1)xSU(2)xSU(3), but we also have B-L, or at least no experiment has been found to violate B-L. So must we match the chiral structure of the gauge groups, or add the B-L to the play? I am not sure if it is the same matching, then.
     
  26. Feb 2, 2010 #25

    Physics Monkey

    User Avatar
    Science Advisor
    Homework Helper

    Hmmm, I think we are pretty sure, right? Roughly speaking, from at least a few hundred GeV on down the world is pretty much described by the chiral gauge theory we all know and love. The fact that at even lower energies the chiral nature of the theory becomes arguably less important is, in my opinion, an unrelated issue.

    Therefore what we should match to is the full standard model as we know it today modulo experimentally acceptable corrections from higher dimension operators.
     
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook