String theory gets you any result you want

[g.lemaitre]

It's really difficult as a layman trying to find out who is right and wrong in this debate as far as string theory. I find it kind of hard to believe that string theory is just simply completely wrong and has no relevance whatsoever to reality. After all so many theorists and so many papers have been written about it, it just seems a bit far-fetched that so many smart people could be heading down a blind-alley. It's much more likely that they have uncovered bits and pieces of the truth. I've read quite a lot of pro-string theory books: Greene, Susskind, Kaku, Hawking and I've only read one anti-string theory book: Woit. There is one argument by Woit that I find very disturbing:

The possible existence of, say, 10^500 consistent different vacuum states for superstring theory probably destroys the hope of using the theory to predict anything. If one chooses among this large set just those states whose properties agree with present experimental observations, it is likely there still will be such a large number of these that one can get just about whatever value one wants for the results of any new observation. If this is the case, the theory can never predict anything and can never be falsified. This is sometimes known as the “Alice’s Restaurant problem,” from Arlo Guthrie’s famous refrain, “You can get anything you want at Alice’s Restaurant.

I think this is a very serious objection. If you've come up with a description of reality that can describe anything then there's no way that I can prove you wrong. It's almost like they have a theory a bit like Max Tegmark's dictum: everything exists. A theory that posits everything exists and that which we cannot observe is in principal unobservable but exists nevertheless I find very dubious.

 

[atyy]

Woit is wrong. Before string theory there were an uncountably infinite number of possible theories of quantum gravity - that is one of the problems of gravity being perturbatively non-renormalizable. 10^500 is a countably finite number.

Eg. Distler says, regarding the pre-string effective theory of gravity: "Then all of these infinite number of coupling become equally important, and we lose control, both computationally and conceptually."

 

[tom.stoer]

Let's make a simple example: Suppose somebody gives you the Lagrangian of the standard model, nothing else. Suppose you are able to construct its (infinitly many) ground states, e.g. the hydrogen atom, water, iron, a neutron star ... that would be a big success!

Suppose now that for some reason all these solutions are not experimentally accessable to you; or that you may speculate to live in one of these solutions but that (up to now) you were not able to identify it exactly. Does that mean that the standard model has lost its predictive power?

There are other reasons why I am afraid that string theory (in its present fashion) is a dead end: it seems that it does not have or allow for a fundamental definition, a small number of defining equations and principles.

Suppose you do not know QED and QM but you only have some approximate rules to combine different atoms (as elementary particles!) to larger molecules and solids. Suppose you have some approximate methods to study things like heat capacity and transport, light propagation, electric conductivity, viscosity etc. Somehow your results are reasonable, they are rather similar to what you observe around you. But instead of identifying QED with its fundamental d.o.f. and being able to "explain" the above mentioned approximate rules for solids you only find new rules - day by day. At the same time you miss liquids, gases ... they are not covered by your rules.

I think that this describes the current status of string theory. There are hints towards some underlying truth but the way string theory is formulated today veiles this truth.

 

[mitchell porter]

There are two questions: is there a string model which completely matches experiment, and, is there a unique or nearly unique string prediction for what observable physics should be like. For both questions, the answer is, we don't know.

Regarding the first question, there are plenty of string models which reduce to something like the standard model at low energies, and you will even see these referred to as e.g. "heterotic standard models". But this just means that at low energies, they will feature the same types and multiplicities of particles, and the same interactions. It doesn't mean that the couplings, masses, or mixings have the same values as they do in the real world; and these quantities are very hard to calculate for any given model. Progress is occurring but it's incremental, e.g. a model is developed in which another feature of reality is present or another non-feature of reality is eliminated. Completely matching experiment, at the level of numerical constants, is still a distant goal; I've heard ten years from now as an optimistic estimate for when the models might get that far.

(Also bear in mind that the LHC is likely to move the goalposts. All these stringy standard models have N=1 supersymmetry, but there is no experimental data on the structure of the supersymmetric sector, if it exists. If the LHC does start turning up squarks and gauginos and so forth, that will have an immediate and positive impact on string model building. If supersymmetry doesn't show up, that has a different set of implications. The measured properties of the Higgs, whether anything, superparticles or something else, shows up as an explanation of how the Higgs mass is so small relative to the unification scale... all of that will affect string phenomenology, the way it is already affecting QFT-based model building.)

Regarding the second question, this is mostly a matter of figuring out how string cosmology should work. Do the initial conditions of the universe lead inevitably to only to one or two special vacua being realized, or do you get an inflationary universe with diverse local vacua which are then anthropically selected, or are there "one-vacuum" universes where the particular late-time vacuum is very contingent on initial conditions? The last scenario would be the least predictive - string theory really would give you a large number of possible worlds and no way to apriori choose between them. The other two are more predictive - either dynamics or anthropic principle singles out special vacua as the ones you are likely to see.

Even if string theory turned out to be "nonpredictive", it would still be a candidate for the description of reality and that would be important; and there might also be qualitatively stringy predictions for astrophysics or cosmology, largely independent of the technical specifics of the vacuum we're in (arising just from the fact that at ultra-high energies, the stringy substructure of particles is supposed to show up).

But all this is very hypothetical, since string cosmology hasn't even settled into one of the three paradigms yet.

 

[Chronos]

How do you dismiss a theory that is so slippery it is inherently unfalsifiable? It will remain a seductive, but, unrealistic path until a manageable limit is set on the direction it should take - which probably explains why alternative approaches have become increasingly popular.

 

[Demystifier]

Let's make a simple example: Suppose somebody gives you the Lagrangian of the standard model, nothing else. Suppose you are able to construct its (infinitly many) ground states, e.g. the hydrogen atom, water, iron, a neutron star ... that would be a big success!

Suppose now that for some reason all these solutions are not experimentally accessable to you; or that you may speculate to live in one of these solutions but that (up to now) you were not able to identify it exactly. Does that mean that the standard model has lost its predictive power?

There are other reasons why I am afraid that string theory (in its present fashion) is a dead end: it seems that it does not have or allow for a fundamental definition, a small number of defining equations and principles.

Suppose you do not know QED and QM but you only have some approximate rules to combine different atoms (as elementary particles!) to larger molecules and solids. Suppose you have some approximate methods to study things like heat capacity and transport, light propagation, electric conductivity, viscosity etc. Somehow your results are reasonable, they are rather similar to what you observe around you. But instead of identifying QED with its fundamental d.o.f. and being able to "explain" the above mentioned approximate rules for solids you only find new rules - day by day. At the same time you miss liquids, gases ... they are not covered by your rules.

I think that this describes the current status of string theory. There are hints towards some underlying truth but the way string theory is formulated today veiles this truth.

This is an excellent description of the current state of string theory.

 

[tom.stoer]

This is an excellent description of the current state of string theory.

thx

 

[negru]

I think this is a very serious objection. If you've come up with a description of reality that can describe anything then there's no way that I can prove you wrong. It's almost like they have a theory a bit like Max Tegmark's dictum: everything exists. A theory that posits everything exists and that which we cannot observe is in principal unobservable but exists nevertheless I find very dubious.

Maybe it's a matter of perspective, but if within that finite number of solutions there are even a bunch relatively close to our SM in the low energy limit, I would call that the greatest success of modern physics. The fact that all SM parameters can be reduced to just one parameter would be the strongest prediction ever produced by physics, and is a highly non-trivial statement.

Some extra random selection principle might exist to restrict the solutions and skip the anthropic multiverse scenario, and that'd be pretty nice, but just a minor and irrelevant detail.

It's sort of like the difference between proving that nature can be modeled accurately by mathematics, and postulating that the speed of light is constant. Which of the previous two statements is more important?

 

[haushofer]

Let's make a simple example: Suppose somebody gives you the Lagrangian of the standard model, nothing else. Suppose you are able to construct its (infinitly many) ground states, e.g. the hydrogen atom, water, iron, a neutron star ... that would be a big success!

Suppose now that for some reason all these solutions are not experimentally accessable to you; or that you may speculate to live in one of these solutions but that (up to now) you were not able to identify it exactly. Does that mean that the standard model has lost its predictive power?

There are other reasons why I am afraid that string theory (in its present fashion) is a dead end: it seems that it does not have or allow for a fundamental definition, a small number of defining equations and principles.

Suppose you do not know QED and QM but you only have some approximate rules to combine different atoms (as elementary particles!) to larger molecules and solids. Suppose you have some approximate methods to study things like heat capacity and transport, light propagation, electric conductivity, viscosity etc. Somehow your results are reasonable, they are rather similar to what you observe around you. But instead of identifying QED with its fundamental d.o.f. and being able to "explain" the above mentioned approximate rules for solids you only find new rules - day by day. At the same time you miss liquids, gases ... they are not covered by your rules.

I think that this describes the current status of string theory. There are hints towards some underlying truth but the way string theory is formulated today veiles this truth.

Apart from experimental access, I can write down an infinite amount of Yang-Mills theories according to the rules of QFT. And nature picks out one of them (one specific product of SU(N)'s, four dimensions, etc.), but we don't understand why. In which sense does this differ from the vacua of ST?

I would say that 10^500 is still smaller than infinity.

 

[tom.stoer]

It differ's b/c you are comparing 'models' with 'solutions of models'. There's a category error.

If you claim that string theory is just a framework like 'gauge theory' to construct models than this may work, but if you want to compare 'string theory' with a specific gauge theory than it becomes vague.

An additional problem is that we are not able to write down a unique and fundamental definition of string theory.

 

[haushofer]

It differ's b/c you are comparing 'models' with 'solutions of models'. There's a category error.

If you claim that string theory is just a framework like 'gauge theory' to construct models than this may work, but if you want to compare 'string theory' with a specific gauge theory than it becomes vague.

An additional problem is that we are not able to write down a unique and fundamental definition of string theory.

Yes. The name theory is a misnomer, it should be " stringparadigm" or something like that. One specific vacuum should be regarded as a theory imo. And that makes the comparison well defined.

The lack of a solid definition is because it seems like we write st in the wrong variables

 

[haushofer]

So i don't see why people complain about the large amount of different string theories. We have an infinite amount of qft's to choose from,and nobody makes a fuzz about that. I don't see why that would be a " category misstake", and why that should matter.

The only thing to complain about is experimental verification, but we're talking about quantum gravity here. So that also see,s abit unfair. Are there qg-candidates which score better falsification-wise?

 

[tom.stoer]

The lack of a solid definition is because it seems like we write st in the wrong variables

Yes!

I don't see why that would be a " category misstake", and why that should matter.

A category mistake simply means that you compare things which cannot or must not be compared. Is darkness blacker than coal?

Are there qg-candidates which score better falsification-wise?

No

 

[audioloop]

I find very dubious.

who knows...



but i agree.
same thing for Loop Theory.

 

[haushofer]

A category mistake simply means that you compare things which cannot or must not be compared. Is darkness blacker than coal?

Sure, but do you agree that if we regard string theory as a framework, there is no problem as such? I regard every particular construction of a string theory (a vacuum of the string framework) as a theory, like the standard model is a particular construction of the QFT-framework. As such I don't see a problem.

On top of that, doesn't the QFT-framework also contain things like instantons and magnetic monopoles which haven't been observed? Do we understand why?

 

[haushofer]

How do you dismiss a theory that is so slippery it is inherently unfalsifiable?

Why is ST "inherently unfalsifiable"?

 

[tom.stoer]

Sure, but do you agree that if we regard string theory as a framework, there is no problem as such?

The only problem left is that it hasn't been completed; it's still under construction. There is no small set of defining equations and rules that can be called 'string theory'. But let's assume that we will find such a definition: then having a huge set of solutions to one theory is of course a major progress compared to a situation with a huge set of theories (with a huge set of solutions)

On top of that, doesn't the QFT-framework also contain things like instantons and magnetic monopoles which haven't been observed?

Instantons cannot be observed when constructing theta-vacua; but instantons (and merons) seem to play an interesting part when it comes to color confinement; so they are definately not unphysical. Regarding magnetic monopoles ... ?

 

[QuantumKitty]

"I've only read one anti-string theory book: Woit."

I suggest reading The Trouble With Physics by Lee Smolin. I found it quite interesting.

 

[Pesj]

Will read Smolin´s book. thanks QantumKitty.