I know the main points of string theory, but can someone tell me where

[HarryDaniels]

I know the main points of string theory, but can someone tell me where the evidence is that it could be true?

 

[tom.stoer]

Warning: I know that other people here will disagree, but as far as I can see there is no evidence.

 

[torquil]

There is no real evidence for this.

Torquil

 

[arivero]

Mathematical coincidences, mostly between branches of string theory itself, but more important also with other independent approaches, such as supergravity and non commutative geometry.

 

[arivero]

I think that this article,
http://dx.doi.org/10.1016/0370-2693(71)90028-1
which elaborates a remark at the end of
http://link.aps.org/doi/10.1103/PhysRevD.4.1109
was more near of the truth that the current interpretation of string theory.

Note that the author is one of the fathers of string theory, but his posture evolved in 1974 http://dx.doi.org/10.1016/0550-3213(74)90010-8 to incorporate the Planck scale.

For this interpretation article, we actually have evidence:

- the pion has the same mass that a charged fermion.
- the total number of "terminated gluons", ie mesons and diquarks, of a given charge is exactly two times the number of fermions in the standard model having this same charge.

 

[tom.stoer]

@HarryDaniels: talking about "evidence" - what exactly do you have in mind? do you have some examples?

 

[HarryDaniels]

@HarryDaniels: talking about "evidence" - what exactly do you have in mind? do you have some examples?

None.
What I mean is: Is there any evidence apart from the fact that it fits in combining the two main theories. Is there any observations that agrees with the theory?

 

[marcus]

I know the main points of string theory, but can someone tell me where the evidence is that it could be true?

Warning: I know that other people here will disagree, but as far as I can see there is no evidence.

But the evidence suggesting that superstring/M-theory is NOT how nature works is overwhelming*. For the moment it seems like a waste of time to discuss it. Some of the top researchers appear to be shifting support and interest, all or part, out of string**. There are other interesting research programs*** which have taken shape in the past 10 years.

*after many years no evidence of extra dim, or low-energy supersymmetry, the failure to find a unique version, explosion into over 10⁵⁰⁰ different versions of physics, politely called the "string landscape". Anti-deSitter (AdS) string preference, against the current consensus of cosmologists, incompatibility of extra dimensions with both dark energy and cosmic inflation according to theorems proved by Princeton's Paul Steinhardt. M-theory still a dream---not yet gelled into definite theory with explicit equations and principles.

**top people such as: Petr Horava, Erik Verlinde, Nima Arkani-Hamed, Hermann Nicolai, Steven Weinberg, Edward Witten.

***other interesting programs, with no spatial extra dimensions, such as: Horava's 4D gravity, Verlinde's 4D gravity, Weinberg AsymptoticSafe 4D gravity, Witten's 3D gravity toy model, Nicolai's program for minimalist non-string unification out to Planck scale, Shaposhnikov's minimalist unification, Loop and spin foam gravity, Path integral 4D approaches such as the Utrecht group's, Noncommutative geometry and noncommutative field theory.

 

[atyy]

None.
What I mean is: Is there any evidence apart from the fact that it fits in combining the two main theories. Is there any observations that agrees with the theory?

String theory is the only known consistent quantum theory of gravity that yields classical gravity described by the Einstein field equations. Whether it describes our universe is unknown. Suggestions for its experimental implications are in eg. http://arxiv.org/abs/1001.0577 , http://arxiv.org/abs/1001.4084 .

 

[marcus]

None.
What I mean is: Is there any evidence apart from the fact that it fits in combining the two main theories. Is there any observations that agrees with the theory?

A major difficulty is, as Gertrude Stein said about the city of Oakland, "there is no there there."

There are many versions of the various string theories. It seems to be difficult to construct one that entirely agrees with all past experimental observation, and such constructions have become increasingly "Baroque" over the years. (Baroque was Hermann Nicolai's word for it. He is a prominent and influential string theorist who is among those who have shifted interest over into simpler "minimalist" unification schemes, which actually make predictions and are testable at accessible energies. As examples of increasingly Baroque elaboration, Nicolai could well have been thinking of "F-theory" papers such as those Atyy just cited in the previous post.)

Moreover, mere agreement with past observation (even if they could achieve it) would not yet be a sufficient test. A theory has to be falsifiable, which means not only agreeing with past observation but also making definite predictions about future outcomes. To gain credibility it has to "bet its life" on specific outcomes of specific future experiments. Some of the new minimalist unification approaches do this.
Nicolai's, for instance, is falsifiable at LHC energies, if the LHC gets up to design specs.

So at this point the whole string enterprise has become of questionable merit. There is a framework which produces a huge variety of different versions. There is no unique theory and there are no unique predictions about the outcome of future experiment. Some of the most eminent people are finding other things to do, devising other approaches to gravity and unification to be interested in---typically 4D approaches. And the trend is born out by sociological measures such as the declining rate of research citations to recent string papers (a measure of value as seen by the researchers themselves.)

So you asked about evidence. I would say that there are many sorts of evidence which suggest that the overall framework or approach (not yet a definite theory) is seen as less promising now than it was, say 10 years ago. There are more exciting things to learn about now, and more exciting developments to watch.

 

[atyy]

There are many versions of the various string theories. It seems to be difficult to construct one that entirely agrees with all past experimental observation, and such constructions have become increasingly "Baroque" over the years. (Baroque was Hermann Nicolai's word for it. He is a prominent and influential string theorist who is among those who have shifted interest over into simpler "minimalist" unification schemes, which actually make predictions and are testable at accessible energies.)

Nicolai and Meissner's work is motivated by string theory. Like F-theory, it uses string theory to constrain suggestions of phenomena that may be observed at the LHC. "However, apart from the known difficulties with (Weyl)² theories of gravity, the known ansaetze at unification in general do not give rise to effectively Weyl invariant low energy theories², despite the ubiquity of dilaton-like fields in supergravity and superstring theory. For this reason we here suggest a different route by exploring whether and under what circumstances it may be possible to get a classically conformal theory out of non-conformal Einstein gravity or some of its supersymmetric extensions." http://arxiv.org/abs/0907.3298"

 

[humanino]

For the moment it seems like a waste of time to discuss it

I do not think this is fair. As much as I have interest in non-string alternatives, I remain convinced that string theory is very fruitful. I wanted to post a historical reference in the lines of what arivero provided. Here is an up-to-date review of string models for gauge theories :

"From Gauge-String Duality to Strong Interactions: A Pedestrian’s Guide"
Annu. Rev. Nucl. Part. Sci. 2009. 59:145–68

It is a fact we should not overlook : there is no comparison out there on the market when it comes to calculating gauge theories in their strong sector, non-perturbative effective methods. I want to emphasize "effective" for two reasons at this level. First reason is that I am interested in results, concrete predictions, and you will not find anything comparing to the gauge-string duality in efficiency when you look into non-perturbative QCD. Secondly, and from this perspective, whether string theory is really fundamental needs not be decided yet. The advances in pure mathematics string theory gave us justify that we still pursue its development.

The last year in particular has seen a lot of publications in effective models, calculations which do work, from string theory. Apart from non-perturbative QCD, I will also mention superconductivity : it is not inconceivable that string theory will provide us with important insights on High-T superconductors.

I would say that there are many sorts of evidence which suggest that the overall framework or approach (not yet a definite theory) is seen as less promising now than it was, say 10 years ago. There are more exciting things to learn about now, and more exciting developments to watch.

There are very interesting, important and fascinating developments outside string theory, and this is great. String theory also made progress during the last 10 years, and it offers many more promises, one can not seriously deny that. Whether other approaches have made more progress, it is possible, but whether that affects the "merit" of string theory, I doubt.

 

[marcus]

It's certainly true that string mathematics is proving useful as a computational tool in fields like superconductivity and nuclear physics. This is where the math is not being used as a microscopic theory of nature but as method to analyze processes at much larger scale. Hermann Nicolai has drawn attention to this in several articles.

Steven Weinberg has spoken recently of string theory as "disappointing" despite his high expectations for it some years back. This "disappointment" seems to be primarily in the area of fundamental particle physics, unification, the "ToE" quest in other words. Here's the video of his October 2009 talk:
http://www.ustream.tv/recorded/2384517

In this perspective talk on the status and expectations of high energy physics, to the 2009 national science-writer's convention, he did not even mention string theory. Not a word about it until someone asked him a question at the end. Here's a transcript of his reply:
http://www.math.columbia.edu/~woit/wordpress/?p=2400
A sample of Weinberg's response to the questioner:

==quote==
It’s developed mathematically, but not to the point where there is anyone theory, or to the point that even if we had one theory we would know how to do calculations to predict things like the mass of the electron, or the masses of the quarks. So, I would say, although there has been theoretical progress it’s been, I find it disappointing...

One of the troubles with superstring theory is that although in a sense the theorists think there is only one theory, there are an infinite number of approximate solutions of it and we don’t know which one corresponds to our world...
==endquote==

What we're talking about is a not-so-subtle shift of emphasis and perspective. Of course we can't tell if interest and activity in string will continue to decline! Something could happen t bring it back into the limelight.

However this process goes, it is interesting to watch and a lot can be told by studying the lineup of talks and events at the annual Strings conference. Strings 2010 is coming up in hardly more than six weeks. It will be at the Texas A and M Campus, at College Station in Texas. The speaker's list has been posted----a great lineup of about 40 speakers. But the titles of the talks have not been posted. It looks to me like every possible effort is being made to re-invigorate the field. It's going to be really interesting to see the titles of the 40-some talks! Here is the website.
http://mitchell.physics.tamu.edu/Conference/string2010/
Steven Weinberg will be one of the 40 invited speakers.

 

[tom.stoer]

String theory is the only known consistent quantum theory of gravity that yields classical gravity described by the Einstein field equations.

Is there a proof regarding consistency?

 

[arivero]

Is there a proof regarding consistency?

Well, it depends. According to v. Neumann, consistency of a set of axioms is equivalent to the existence of a model where these axioms evaluate to true. Except by exhibition of a model I am not sure if there is other way to prove consistency.

What is usually understood here is absence of anomalies, divergences and all that stuff.

 

[tom.stoer]

What is usually understood here is absence of anomalies, divergences and all that stuff.

I know.

But as far as I know there is not even a definition for the n-loop amplitude in superstring theory, not to mention a proof of its finiteness.

 

[arivero]

I know.

But as far as I know there is not even a definition for the n-loop amplitude in superstring theory, not to mention a proof of its finiteness.

Well, God does not make perturbation theory, they say.

 

[tom.stoer]

So can you ask him for a non-perturbative quantization of superstring theory?

 

[torquil]

My impression is that string theory at the moment is "less" well-defined than QFT. But I think mathematicians say that QFT is not completely well-defined either. Of course, QFT has an impressive history of achievements.

Torquil

 

[tom.stoer]

To make this a precise statement: There are two reasons to believe in a theory:
1) it reproduces / post-dicts known facts and predicts (correctly!) new phenomena
2) it looks like a fundamantal, consistent and "appealing" mathematical concept
Of course 2) use useless w/o 1)

1) is true for quantum mechanics, the standard model etc.
1) and 2) is true for general relativity
Of course for 2) it's somehow a matter of taste

String theory fails according to 1) It does neither post-dict well-known facts (it failes to reproduce the standard model, but it comes close to it) and it makes no new predictions (there are predictions which have to be hidden, e.g. 10 dim space-time)

So we are left with 2) Here the claim is that string theory overcomes the usual difficulties of quantum field theory in the sense that it is manifestly finite = free of divergences. Unfortunately there is no proof!
a) there is no fully understood non-perturbative quantization scheme
b) there is no definition of a perturbative quantization scheme beyond a few (2) loops. I studied a paper regarding the 3-and 4-loop amplitude a few month ago, but the results seemed to be incomplete). There is no proof of its finiteness, either.

So we are left with the assurance that in the furture (when ?) 1) will turn out to be true; and we are left with the claim that string theory is "appealing" according 2). Unfortunately facts and proofs are missing.

I am no expert in string theory. What I see is that they are working rather hard to make progress with 1), especially the F-theory approach comes quite close to the MSSM. But I am missing results from 2)

 

[atyy]

My impression is that string theory at the moment is "less" well-defined than QFT. But I think mathematicians say that QFT is not completely well-defined either. Of course, QFT has an impressive history of achievements.

To make this a precise statement: There are two reasons to believe in a theory:
1) it reproduces / post-dicts known facts and predicts (correctly!) new phenomena
2) it looks like a fundamantal, consistent and "appealing" mathematical concept
Of course 2) use useless w/o 1)

1) is true for quantum mechanics, the standard model etc.
1) and 2) is true for general relativity
Of course for 2) it's somehow a matter of taste

String theory fails according to 1) It does neither post-dict well-known facts (it failes to reproduce the standard model, but it comes close to it) and it makes no new predictions (there are predictions which have to be hidden, e.g. 10 dim space-time)

So we are left with 2) Here the claim is that string theory overcomes the usual difficulties of quantum field theory in the sense that it is manifestly finite = free of divergences. Unfortunately there is no proof!
a) there is no fully understood non-perturbative quantization scheme
b) there is no definition of a perturbative quantization scheme beyond a few (2) loops. I studied a paper regarding the 3-and 4-loop amplitude a few month ago, but the results seemed to be incomplete). There is no proof of its finiteness, either.

So we are left with the assurance that in the furture (when ?) 1) will turn out to be true; and we are left with the claim that string theory is "appealing" according 2). Unfortunately facts and proofs are missing.

I am no expert in string theory. What I see is that they are working rather hard to make progress with 1), especially the F-theory approach comes quite close to the MSSM. But I am missing results from 2)

Would string theory be "less" well defined than QFT if it had had experimental success already?

 

[humanino]

especially the F-theory approach comes quite close to the MSSM.

There are so many attempts, to my knowledge this is the best example of "proof by exhaustion of the audience". I would however like to emphasize, at this point it is not a problem of whether string theory can reproduce low energy physics. It certainly can, in many ways. The problem is that doing so requires quite some tuning, choices must be made which eventually do not allow for a satisfactory "explanation". So, there is no sense of "uniqueness", but I doubt one can claim that "the MSSM cannot be reproduced" : there are simply too many articles and PhD thesis out there claiming to do so, and one would have to spend their entire life refuting them.

As for predictions in string theory, one could argue, perhaps more convincingly, that supersymmetry and additional dimensions are required. In any case, they are required in the vast majority of string models, even if one can cook more "exotic" solutions (or less explored parts of the landscape) with 4D or no supersymmetry on the world-sheet.

 

[arivero]

A part of the history of string theory is to use the lack of fit with the standard model to justify the abandon of an area and the rise of another. Of course they move from an area of research to another for a veriety of reasons, mostly an hibrid of sociology and productivity.

As for F-theory, I find its adoption intriguing. It was known since Bailin and Love works that a full reproduction, in Kaluza Klein, of the standard model charges required to use 8 extra dimensions. I think -but I can be wrong- that the intuition is that you need to have an extra U(1) for the B-L charge, which appears both in Weinberg model and in Pati Salam, but it need to be fully broken. Again intuitively, full broken amounts to infinitesimal extra dimension (or, infinite mass of the carrier). So the insight with kaluza klein fits with the F-Theory.

And it should not. Because string theory, and with it F-theory, does not get all the gauge groups from space-time.

 

[atyy]

So, there is no sense of "uniqueness", but I doubt one can claim that "the MSSM cannot be reproduced" : there are simply too many articles and PhD thesis out there claiming to do so, and one would have to spend their entire life refuting them.

In Zwiebach's text (I'm not sure which edition) he says that string theory has a problem with the Higgs, and as of 2007, Distler said "If we should be disturbed by anything, it is that, out of the plethora of string vacua found to date, none of them looks sufficiently like our world, rather than that there are too many that do." http://golem.ph.utexas.edu/~distler/blog/archives/001200.html Has the situation changed since then, or do you think they disagreed that with those works claiming to come close enough to the MSSM?

 

[Haelfix]

There are a lot of dead ends in stringy phenomenology where you can get a lot or most of the features of the standard model, and then for one reason or the other the rest is either wrong or (more commonly) out of calculational control. Situations where you can say "in principle this might be the vacua of the real world, but we simply can't reliably compute what this superpotential tells us for say the low energy values of the CKM matrix". Often we are talking about ridiculously difficult calculations, that are outside of analytic or even approximate control.

So a lot of model building focuses on schemes where such calculations are possible as opposed to the most aesthetically pleasing and simpler ones. But then nature doesn't really care what is or is not technically hard for humans either.

The veil is thick in general with stringy model building and the real trick is figuring out where to even look before wading into a multiyear calculation.

 

[arivero]

The veil is thick in general with stringy model building and the real trick is figuring out where to even look before wading into a multiyear calculation.

Yes, and here enters "sociology": You look under the lamp instead of looking of the darkness. Meaning, you look on calculations which are small variants of the calculations your group is trained on. In this way, different universities/research groups wade across the "landscape". Of course the other part of the argument is around productivity and citations: if you are to choose between some different options available to your group, the one with more impact to other groups is to be chosen.

Regretly, all of it means that an abandoned area of exploration is not easily populated, even if the logical reasons to abandon it do not exist anymore.

And fortunately, it is not at all negative; even a random slime-mould way exploration of the terrain will be affected by the nutrients in this terrain. So the main features string theoretist are looking at (supersymmetry, 10D, dualities, M-theory, F-theory, etc) are probably related to the theory of the world. But not forcefully in the expected way.

 

[arivero]

Just to put an example: the M-theoretic duality between heterotic and type A theories. It is usually presented as a duality between compactification in a segment $$I=S^1/SO(1)$$ and a circle $$S^1$$. But on the same token, you could call these spaces $$S^3/SO(3)$$ and $$(S^2 \times S^1)/SO(3)$$. So M-theoretical duality translates between an space with symmetry $$SO(4)=SU(2) \times SU(2)$$ and another one with symmetry $$SU(2) \times U(1)$$. And it could be said that string-theoretists have located, in his abstract random way, a phenomena well known of the generalisations of the standard model: the relationship between R-L models and the electroweak model.

 

[tom.stoer]

A certainly agree that string theory is far from successful low-energy phenomenology and model building. I agree that string theory has a lot of dead ends, that it requires fine-tuning w/o direct explanatory statements. I am not a fan of string theory - basically because of all these shortcomings.

But what I appreciate is that string theorists started to work on these topics!

Some years ago the criticism was that they do not talk about physical reality; now they really concentrate on these issues and we should let them continue.

My opinion is that they will not succeed - but everybody is free to disagree!

 

[Fra]

it is not a problem of whether string theory can reproduce low energy physics. It certainly can, in many ways. The problem is that doing so requires quite some tuning, choices must be made which eventually do not allow for a satisfactory "explanation". So, there is no sense of "uniqueness"

This is my main objection to string theory as well. The thing with compactified higher dimensions etc is not really a "problem" aside with the apparent ambigousness of the class of theories.

This is why my only personal hope/motivation for string theory, is in a context where string are approximately emergent continuum structures from a discrete framework (where the low complexity limit should be unique; as opposed to systems with not only infinite but also uncountable complexions - then strings, branes and other things are emergent by stability as the complexity is scaled up). In this sense the duality between the various dualities in string theory can probably be more natural too, that two apparently different microstructurs can still encode the same information as measured by their interactions.

But if anything like this is every worked out, I think it will still be a fundamentally different construction that string theory based on different principles.

/Fredrik

 

[arivero]

I really do not believe that string theory can reproduce the standard model in many ways. They use a kind of sylogism: "Mathematics can reproduce the standard model in many ways", "String theory can produce a lot of Mathematics", thus "String theory can reproduce the standard model in many ways". The first part is false.

It is easier to believe that there is an unique or essentially unique way to produce the standard model, and that string theory can produce it. A lot of uniqueness is already in, and it is well understood that it relates to mathematical uniqueness: Evans shows that existence of supersymmetry and membranes is linked to the existence of division algebras. The source of non uniqueness in current research is the vacuum selection, which allows them to build arbitrary SU(N)^M groups even when the internal SO(32) is not available. My guess is that they are confusing gauge and flavour groups.