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

A Theory We're Incapable to Understand?

  1. Jun 4, 2010 #1
    Since the 1960s (the dawn of String Theory) Theoretical Physicists have been examining the complex structure of one of the most 'profound' theories to be developed by Mankind - String Theory. In 50 years the Theoretical Physicists working on this field have developed it further and have been able to understand the various perplexing ideas that underly this theory. As everyone would know in 1995 the theory was drastically redefined and a new spark of interest was ignited. It has been 15 years since the creation of M-theory and my questions are: Is M-Theory still in a state of obscurity? Do we understand how it relates to the physical world? Has there been any paradigm shift in how we see this theory? And finally, what dilemmas does this theory present that makes it so difficult to understand?
  2. jcsd
  3. Jun 5, 2010 #2
    The question you ask here reappers permanently on the forum. Perhaps you find some answer in "surfing" on equivalent pages... Otherwise -and it is only a very personal view- I cannot believe that specialists have worked 50 years long without being able to bring some progresses... even if only for mathematics.
  4. Jun 5, 2010 #3


    User Avatar
    Science Advisor


    I agree that you may find other threads where these questions have already been discussed. And let me warn you that you will hardly find - neither here nor elsewhere - unbiased opinions regarding string- / M-theory ...

    yes and no.

    We know of certain mechanisms how physical structures like our world (4-dim., the standard model, ...) can emerge from M-theory; but there is neither a construction from which the standard model exactly follows, nor is there an idea why regarding its uniqueness - why it is not something totally different; so there is no idea why the world is at it is (except for the antropic principle which is rather obscure).

    An approach in the context of M-theory which comes rather closed to SM-like theories is F-theory.


    - the second superstring revolution (which you already mentioned)
    - the landscape discussion

    - it misses a fundamental guiding principles
    - it misses a set of defining equations and calculational tools

    In ordinary QFT there are some equations which define the theory; the calculational methods are used to solve it; my impression ist that in string theory the calculational methods are still used to construct the theory.

    Perhaps the situation is comparable to the "old quantum mechanics" before the break through of Heisenberg, Schrödinger and Pauli.
  5. Jun 5, 2010 #4


    User Avatar

    Here is another biased post

    This is my impression too, as I see it the "guiding principles" seems to have been (except for the really early idea that ST was a candidate for strong interaction, which then lost the battle to QCD) largely like what I consider like a simple almost "crap shot" kind of idea, namely that many problems with physics - QG and renormalization in particular - was that we consider POINT particles. Then what if the points particles are really some extended objects that are oscillating (namely STRINGS) that from our observational scale just looks like points, or is indistinguishable from point. There is also a postulated microstructure of the string that comes with an action. This as I see it the "idea" that started it.

    As and idea, there is notthing wrong with exploring it.

    The idea was then to keep all existing QFT methods and framework as is, except try to extend it to get strings in the picture. When that started, it was consistency requirements due to requiring that the extension of QFT methods must gold at all scales and all times that came up with various "predictions" of higher dimensions like "if string theory as postulated is to make sense, we need this or that dimensions". It was quickly noted that the idea of strings, made the divergences alot better! But it was also found that these ideas, gave rise to several different string theories.

    Then as time went on, further ideas along the lines that "if string theory is to make sense..." then then various versions of string theories, that seem related to dualities must be unified in a larger theory M-theory.

    Also there has been the landscape problem all along, that got WORSE when going to M-theory. Which means there are undetermined choices that is needed to define the physical theory. They theory apparently makes no prediction on this.

    As I see it, this is how the ideas has evolve and keeps evolving.

    If something meaningful will come out of this remains to see, but I think if you look at it's development I agree with Tom that it's a strange and weak "motivating guidance".

    Somehow string theory avoids all the deeper question, such as foundations and nature of QM, law, problem of time etc. The idea to keep all the structure of QFT and just add the idea that points are to be replaced by fundamental oscillating strings seems conceptually too simple to solve the problems.

    It would have been nice if it was a quick and dirty temporary fix for renormalizabilty, but as now decades has passed, it seems it was not so easy after all. Make that's a hint towards the "if string theory is to make sense..." ;)

  6. Jun 5, 2010 #5


    User Avatar
    Science Advisor

    I was reading a paper last year where they discussed how to define 3- and 4-loops in ST. That means they do not have an expression for the n-loop amplitude in superspace; how shall they then prove its finiteness? not to speak about the convergence of the series? => if ST should make sense then only non-perturbatively.

    I am not sure if ST must ask the "deep questions" regarding quantum theory. If you look at other approaches towards quantum gravity this questions seems not to be necesary.

    There is one idea which bothers me for for months: many ideas, dualities etc. in ST are based on large-N relations. What if ST is a large-N approximation to some underlying field theory?
    Last edited: Jun 5, 2010
  7. Jun 5, 2010 #6


    User Avatar

    This is also a matter of bias I guess. It's true that many other approaches, LQG for example do NOT ask deeper questions, but then again, they are not satisfactory either.

    Maybe I just happen to a bit more radical due to the way I see things. For me neither LQG nor ST ask the right questions, but I still can find interesting ingredients in both.

    I believe more firmly that the information theoretic, inference and evolution based approaches are more rational. This is so much more immature though, that it's not possible to compare with somewhat "older" disciplines like ST or LQG. It's like comparing apples to seeds.

  8. Jun 5, 2010 #7
    Thanks, I've been on this website frequently for the past six months and I've been constantly reading this part of the forum. I've looked at all of the threads and none of the answers seemed satisfying so I guess I didn't look throughly, much appreciated for you 'guys' to take the time to answer though.
  9. Jun 5, 2010 #8


    User Avatar
    Science Advisor

    You will not find many answer from insiders here - mostly outsiders; so perhaps you should ask the same questions to string theoriests.
  10. Jun 5, 2010 #9
    To: tom.stoer
    Yea, I live fairly close to the Perimeter Institue - approximately fourty-five minutes away, Stephen Hawking is actually arriving today I believe. I'm only sixteen so it was just a strain of curiosity, my understanding is extremely pre-mature and I know I have a long way to go in my level of knowledge. This is something I wish to pursue (well theoretical High Energy Physics) and I enjoy learning so I thought I would ask some questions I have had. So thanks, and if I see a String Theorist I will have my questions prepared.
    Last edited: Jun 5, 2010
  11. Jun 5, 2010 #10


    User Avatar
    Science Advisor

    good luck!
  12. Jun 5, 2010 #11
    So your professional interests are in LQG research?
  13. Jun 5, 2010 #12
    I'm quite certain he stated that he doesn't believe any theory as of now is a plausible candidate for a theory of everything which includes LQG and String Theory. I think he works in Condensed Matter because I thought I saw him mention this in another thread.
  14. Jun 5, 2010 #13
    Sorry, I have one more question.
    On Wikipedia it states under the section of M-Theory that in late 2007 an equation was developed specifically, the Bagger-Lambert-Gustavsson action. If you have any knowledge of how this provides a "long-sought microscopic description of M-Theory" can you explain to me how? If it's far too technical to explain in simplistic terminology then just ignore me because my knowledge is too primitive to understand the mechanics.
  15. Jun 6, 2010 #14


    User Avatar
    Science Advisor


    I was aktive in quantum field theory topics: topological methods and canonical quantization in QCD. Ashtekar's formulation came out one our two years before we started to work on something very similar in QCD, but we were not aware of his approach. Then I had the idea to do something similar in quantum gravity (which I knew from a talk about the Wheeler-deWitt equation). But then I found a book in which Ashtekar described his approach. So everything I had in mind had already been done.

    As canonical quantization in LQG is formally very closed to gauge theory I was always studying their papers and I had contact to some people in the field.

    Today I am no longer active in physics but working in process automation / SW development. I try to keep up with certain subjects (QFT, QCD, LQG) even if it's rather hard if you are no longer an insider. From time to time I try to listen to Thiemann's talks (he is teaching at Erlangen, Germany which is quite close my home).
  16. Jun 6, 2010 #15


    User Avatar
    Science Advisor

    LQG as of now is no candidate for a ToE by construction. The LQG approach is restricted to quantum gravity. It is compatible with matter interaction, the mathamtical formulation (gauge theory) is rather close in a certain sense. But there are no direct attempts to unify gravity with matter.

    There are some ideas how particles could emergy from "braided" or "twisted" spin networks, but that seems to be highly speculative. If you like I can give you some references.

    There is the idea to harmonize non-commutative geometry with the LQG approach which could lead to standad model matter emerging in the LQG framework; again this is an idea, not a fully developed research program. You can get references if you like.

    ST is certainly the candidate for unification. But as I said I do not believe in it.
    Last edited: Jun 6, 2010
  17. Jun 6, 2010 #16


    User Avatar

    Sounds like a nice part of Germany to be in. I like bavaria alot, and if I had to move to germany bavaria would probably be my choice. We are cooperating with plenty of companies in the munich aread so I've been to munich numerous times the last few years. One them are also industrial automation companies (can, profibus and opc technology). Along with belgium, bavaria has the best beer in the world so it's not a bad place to be :)

  18. Jun 6, 2010 #17


    User Avatar
    Science Advisor

    Roughly, our current understanding of M-theory is like understanding of electrodynamics at the level of Faraday, before the Maxwell equations have been discovered.
  19. Jun 6, 2010 #18
    Why is that? Are you skeptical of GUT's, higher dimensions, or SUSY?
  20. Jun 6, 2010 #19


    User Avatar

    There are alot of papers on this, both fomr many string theorists and even connecting back to 't Hooft papers from the 70's on strong interactions.Usually when these questions are phrased in a framework that is part of what I think is the problem, so I don't have any comments on this in the original context as I probably wouldn't be asking the question like that at all.

    But like I've mentioned before, my one possible optimistic view of ST is based on a possible connection to the view I have, and it's rouglhy that the continuum string is a pretty much associated with the simplest continuum object that emerges as one reconstructs discrete measures on discrete index spaces.

    Very simply put, a 1D string embedded in a 2D space, can be associated to a probability distribution. Where the event index, is the string index, and the mass per unit index corresponds to the probability density, or "evidence count per index".

    I also associate this "discrete string" to be like the prototype of a simple observer.

    In this sense one can also infer a rational action of this simple observer = string that is in line with a kind of entropic dynamics = no need to "postulate" string action from classical mechanics analogies of strings - there would be a deeper informaiton theoretic understaning of string action based on information divergence of encoded in the string.

    Now as one consider this in a context of evolving observers, this simple case is complicated, and more complex structure than just "sequences of counts" emerges. It's in this context, I can make sense of the "consistency constraints" that suggests that maximum fitness takes place in special dimensions.

    Clearly information wise, a string in the above sense, can be transformed into different structures, that could in the continuum limit be interpreted as either 1D strings in higher dimensions, OR p-D "branes" in some dimensions, or somehow systems thereof. This is really complex though and I'm still struggling with this.

    In this view, then also one would get a better understanding of the "embedded space" of the string. In this view, this embedding would not be real, two observers can disagree about the embedding space, and objectivity is only in the _relation_ between observers, and this is evolving as observers do.

    So thinking about a string embedded in an higher dimension, is an exernal picture that really isn't right. In the intrinsic picture the embedding is just an expectation of the string itself, that corresponds to a maximally informative and fit representation.

    Structures that fail do realise this, will loose their confidence and mass to it's environment, in a darwinian style.

    Also this view, could give new views on the landscape problem. The landscape problem is just the problem of single out a preferred observer, and this can be done of course, and this is why the question should be posed differently, and instead of inflating an imaginary non-real landscape leaving us with a choices that are undecidable one should reconstruct the very framework of asking questions.

    This is what to me, the discrete reconstruction of the inference is about. It shares view view Ariel Caticha, but it rejects the objectivity of the statistical manifold. My point would be, that each observer really does SEE a difference manifold!

    The only managable starting point I see is the unified perspective. This is why my starting point is unification of forces, and I try to understand how the diversity of symmetries emerge as the complexity ALLOWS SO (as we go down the energy scale from unification).

    To me, the unification energy scale corresponds not at this point to some definite Joule number, it rather corresponds in my view far as "zero complexity".

  21. Jun 6, 2010 #20
    I am not a specialist; so I shall ask more than I can give answers. Was the guiding principle of ST not the following: "A 3D microscopic sphere traveling through time is a kind of string?"
  22. Jun 6, 2010 #21


    User Avatar

    This corresponds that there are two views of unification, the external view and the internal view.

    I argue that the external view, can never make sense, because each observer has a natural complexity cut-off, which is the complexity of the observer itself, so the theoretical "infinite-energy" limit is non-realisable.

    The "external view" usually mathematically imagines an infinite information-sink that can contain an infinite amount of information. But it should be clear that such a picture can never be maximally fit relative to a finite computing device.

    But unfortunately, and this is my main critique towards ST is that to understand this better we need to go deeper than a string. I don't think strings are elementary (if we have tno talk about "strings" at all, which I think we don't). Because if you ask in the unification picture. how does the simplest possible observer (the STRING in string theory) infer the laws of nature? then that depends on how the string is embedded, and I think to understand this you need to understand why there is a string there in the first place.

    The only preferred role of "STRINGS" i see is that they are as far as I can see, the SIMPLEST continuum object! But, they are by no means the simplest object if you don't start with the continuum!

  23. Jun 6, 2010 #22
    Coming back to a trivial reality: one usually thinks strings as oscillating objects (oscillation between two fixed points) but one never envisages the elastic elongating strings. Why?
  24. Jun 6, 2010 #23


    User Avatar
    Science Advisor

    Ordinary GUTs w/o SUSY have been rules out by experiment.

    Breaking SUSY spoils the original elegance of SUSY completely. You always need fancy copies of SUSY and many Higgses in order to achieve that. One introduces something rather simple and elegant, and then work for decades in order to hide it, to break it etc. It doesn't look right.

    Higher dimensions are nice provided that somebody can explain why the world we see is four-dimensional. But they can't neither tell us why it's four-dim. nor why the SM is at it is. So again one introduces a rather elegant structure (bosonic and fermionic ST is really nice!) and and then invests decades in order to break all these nice features. It doesn't look promising after all.

    I will change my mind as soon as somebody presents a simple and clear picture instead of his patchwork.
  25. Jun 6, 2010 #24


    User Avatar
    Science Advisor

    One does something like this, but this "longitudinal mode" is constraint by a symmetry and eliminated via gauge fixing. This is similar to a longitudinal photon which is also eliminated by gauge symmetry.
  26. Jun 6, 2010 #25
    I'd change my mind if LHC (or other observations) finds compelling evidence for these :)
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook