Why Do We Need a Quantum Theory of Gravity?

[marcus]

Why do we need a quantum theory of gravity?

Lots of people have asked. It is a good question to ask because it evokes interesting answers. I will tell you my current answer---we can all have different reasons, and they can change from year to year---there is no single correct answer for all time, to this question.

Most recently, the Amsterdam physicist Theo Nieuwenhuizen ("theon")
http://staff.science.uva.nl/~nieuwenh/
asked this question in a conversation between several physicists at Peter Woit blog Not Even Wrong
http://www.math.columbia.edu/~woit/wordpress/?p=308
there are currently 39 comments and "Theon" comment is number 35
so to find it you can go to the end and scroll back a few. He said:
"Did anybody question whether we really have to quantize gravity?"

The answer is obviously yes, I have seen this questioned several times, and discussed. So people HAVE questioned whether. But probably Theo knows this, and only asked to encourage people to think.

Maybe we at PF can actually use Theo's question better than the people at N.E.W. because, for one thing, the multithread format of PF board is better than the blog format. In blogs like Peter's, a conversations get covered over and lost very quickly. New comments on a thread do not bring the thread back into sight. (Cosmic Variance has a rudimentary sidebar menu of active threads, but most blogs do not.)

 

[marcus]

I had to go to another thread. I didnt yet say why quantize gravity.
My current reason to do this, that I see, is contained in Freidel's latest paper.

He gets Feynman diagrams out of spinfoam. He gets an effective QFT as the "semiclassical limit" (the zero gravity limit as G -> 0) of a spinfoam model of gravity.

This is done rigorously in 3D. Now that stage is finished and was successful---so they are starting to try the same thing in 4D. There is a Freidel/Baratin paper in preparation. Freidel gave the overview at Loops'05 and talked a little about extending to the 4D case, and then Baratin gave a talk at Loops'05 where he presented stuff from the paper they have in preparation.

the paper in prep is called Hidden quantum gravity in Feynman diagrams. The idea is (what is already shown in 3D) that there is a SPECIFIC spinfoam model which you must have if you want it to have feynman diagrams in its flat, zerogravity, limit.
So a particular QG model is HIDDEN already in the flat, or effective, QFT.

We know that current QFT cannot be fundamental because it is built on a flat non-dynamic spacetime----so QFT is only an effective theory which is an approximation (when gravity is negligible) of the real QFT of matter. OK, Freidel says that he can go beyond the effective theory of matter to a more fundamental theory simply by constructing the unique Quantum Gravity model which the flat gravityless QFT insists on---the spinfoam QG which is "hidden" in the Feynman diagrams.

this is a way of saying why it should be useful to quantize gravity. because it will lead to a more fundamental version of the Standard Model

 

[selfAdjoint]

Well, suppose this was true:

• GR is correct, as interpreted with equivalence classes of diffeomorphisms, but otherwise classical "all the way down". No gravitons, no quantized spacetime; gravity is curved geometry.
• Some extension of the Standard model turns out to work in curved geometry
• And some classical interpretation of quantum measurement turns out to be workable.

Would that be so bad?

 

[Stingray]

Well, suppose this was true:

• GR is correct, as interpreted with equivalence classes of diffeomorphisms, but otherwise classical "all the way down". No gravitons, no quantized spacetime; gravity is curved geometry.
• Some extension of the Standard model turns out to work in curved geometry
• And some classical interpretation of quantum measurement turns out to be workable.

Would that be so bad?

Even if spacetime remains completely classical, what determines it? Einstein's equation uses the classical stress-energy tensor as a source. This loses its meaning if matter is described quantum mechanically (as it apparently must be). Simply replacing the right-hand side of Einstein's equation with the expectation value of the stress-energy tensor leads to inconsistencies. So GR must change as far as I can see.

 

[marcus]

Well, suppose this was true:

• Some extension of the Standard model turns out to work in curved geometry
  

Would that be so bad?

devil's advocate

and when there is an extension of Std Mdl that works on curved geometry, then what curved geometry do you choose for it to work on?

I actually don't think we have to discuss whether we can MAKE DO with a schizophrenic mix of classical Gen Rel and extensions of the Std Mdl.

this question of "needing" QG is not a practical question. Squirrels live happily without physics and humans can be happy too. We don't need QG for survival but as Fabien Besnard said "for the honor of the human mind"

So this question of Theo in Amsterdam is like a divergent series. Divergent series are good to consider because they give us ideas. hardy said "this series diverges, therefore we can use it to derive numbers!"
THERE IS NO ONE CORRECT ANSWER to why we must find QG.

today, my answer is to point at the work of Laurent Freidel. He has found a spinfoam QG, with matter, which CONTAINS feynman diagrams and reproduces vertex amplitudes. So QG offers a way to a more basic theory of matter (as a zerogravity limit of a theory of spacetime-and-matter)

the latest Freidel is great---I should get a link in case anyone hasnt seen it.

here are some PF posts talking about the Freidel work and giving links
https://www.physicsforums.com/showthread.php?p=856328#post856328

https://www.physicsforums.com/showthread.php?p=856369#post856369

 

[garrett]

Stingray's simple explanation of why gravity needs to be quantized is correct.

 

[Locrian]

Hmm.

The only reason to quantize gravity is to match experimental observations that can't be produced with theory, or invalidate current theory.

Which ones do that?

 

[selfAdjoint]

Even if spacetime remains completely classical, what determines it? Einstein's equation uses the classical stress-energy tensor as a source. This loses its meaning if matter is described quantum mechanically (as it apparently must be). Simply replacing the right-hand side of Einstein's equation with the expectation value of the stress-energy tensor leads to inconsistencies. So GR must change as far as I can see.

Well, I sort of addressed that in my third bullet point. See Paul Merriam's paper at http://www.arxiv.org/abs/quant-ph?0506228. Maybe a relative approach gets us a "classical" system.

 

[turbo]

Well, suppose this was true:

• GR is correct, as interpreted with equivalence classes of diffeomorphisms, but otherwise classical "all the way down". No gravitons, no quantized spacetime; gravity is curved geometry.
• Some extension of the Standard model turns out to work in curved geometry
• And some classical interpretation of quantum measurement turns out to be workable.

Would that be so bad?

What if:

• GR is incomplete and curved geometry is only an approximation for gravitation. The correcting extension of GR will entail a local mechanism for gravitation and inertia that can be explained by matter's interaction with a single field.
• The standard model requires at least one modification (i.e. no Higgs boson) but it otherwise correct.
• The single field in option 1 is the quantum vacuum field. (Sakharov's conjecture)

Such a scenario would retain the standard model with its many successes and incorporation of the field of the quantum vacuum into GR as the source of gravitation and inertia may entirely remove the need for DM and DE.

 

[Chronos]

Isn't any model necessarily 'flat' within it's own reference frame? If you push any mathematical model through enough integrations, does it not ultimately flatten out?

Conversely, is there really any point in mathematically modeling things beyond observational reality? Do those models necessarily have meaning? I think not. There are many mathematical constructs that have no observational consequences. They might be beautiful, but, still irrelevant IMO.

So, in other words, I agree with Stingray.

 

[vanesch]

Well, suppose this was true:

• GR is correct, as interpreted with equivalence classes of diffeomorphisms, but otherwise classical "all the way down". No gravitons, no quantized spacetime; gravity is curved geometry.
• Some extension of the Standard model turns out to work in curved geometry
• And some classical interpretation of quantum measurement turns out to be workable.

Would that be so bad?

I think the main problem is with your 3rd point, IF Bell situations are real. There's no way in which you can have a classical interpretation of quantum measurement that can respect locality (if we accept the Bell type experiments) ; and if you cannot respect locality, GR falls on its flat face. That's why people who cherish GR have some aversion for Bell situations (like our friend Careful).

So or there is some (non-local) collapse of the wavefunction (which will never be able to be formulated in a generally covariant way), or there is NO such collapse, in which case you can have gravitationally significantly different quantum states in superposition, and then the question is: how do you determine a stress-energy tensor from such a superposition ?

Now I have been "dreaming" about a possibility, in which we have a global wavefunction consisting of a "matter" part and a "geometry" part, and you could think of writing:

|psi> = |matter1>|geometry1> + |matter2>|geometry2> + ...

in which you have of course to write the "matter" states in a localized basis (based upon the conjugated geometry state), and which DO have a well-defined energy-momentum tensor. As such, this introduces of course a preferred basis (which looks very classically).

Now, my wet dream would be that in the Schroedinger equation, the "time derivative" becomes a state-dependent operator:

instead of having d |psi> / dt, you would have a DT operator, which reduces to a time derivative wrt to a foliation of each geometry:

DT|psi> = [d/dt1 |matter1>] |geometry1> + [d/dt2 |matter2>] |geometry2> + ...

where d/dt1 only makes sense wrt "geometry1" of course, and d/dt2 only makes sense wrt "geometry2"... and we'd have individual Schroedinger equations for each individual term.

geometry1 could be totally classical (respecting GR), while "matter1" would evolve according to the curved spacetime of geometry 1... UNTIL it becomes gravitationally "superposed" in which case we'd have to split the term into two different terms "matter1a" and "matter1b" with "geometry1a" and "geometry1b".

The difficulty will be to recombine these different terms into one single wavefunction. One might even hope that we get out a non-unitary combination that way (although the individual matter terms do evolve strictly unitary, the recombination into one wavefunction might be non-unitary).

Now, I should send myself warning points for over-speculative stuff :-)

It is rather my question: I'm pretty sure that such a semi-classical approach has already been considered and it must have a lot of problems.

 

[selfAdjoint]

Patrick, do have a look at the paper I linked to in post #8, building on Rovelli's relational approach, Merriam imagines a law that requires different interacting systems to see the same value of h (a la c with relativity) which requires a relative adjustment of their mass scales.

Now imagine in the spirit of DSR that the laws of nature require all systems to see the same values of c, G, and h (and thus in particular the same values of the Planck dimensions), Work out the transformations that do this and then build it into the tangent spaces of, umm, Relative General Relativity (RGR).

 

[Chronos]

A naive question: Is it necessary to break this down to a wave function?

 

[hossi]

Another reason why gravity should be quantized: singularity avoidance. The appearance of singularities seems to tell me that we are doing something wrong. However, though this seems a good motivation for me, it is not so completely strict. I.e. modifications on a classical level (meaning non-quantized but beyond standard) might do that.

I totally agree with Stingrays argument, just ask: what is the gravitational field of a quark? However, also that does not necessarily mean that gravity itself has to be quantized, just that we don't really understand QFT in curved space or its observables. BTW, intersting paper about that

http://xxx.lanl.gov/abs/hep-th/0512200"
Authors: Steven B. Giddings, Donald Marolf, James B. Hartle

What I mean is: if we understand how the source-term arises from a quantum theory (right hand side) that does not imply that we have to quantize gravity (left hand side).

However, what is clear is that gravity at the Planck scale is a key point for our understanding of a unified theory. And therefore pretty addictive - at least for me.

What if instead of quantizing gavity, gravity is the CAUSE for quantization?

 

[marcus]

What if ... gravity is the CAUSE for quantization?

simply want to echo that question

 

[Careful]

Hi,

Let me first say that this is a good question and that many physicists actually see no need at all for this quest. Most take the pragmatic attitude that it is sufficient to have generation C of the standard model + Einstein gravity (with the *correct* cosmological model) to calculate whatever you desire. For me, QG is about the clash between different mutually exclusive religions, so clearly at least one has to be sacrified (it is fashionable to sacrifice the superior *theory* - which is GR). Therefore, QG is the name given to the Platonic desire to know how two contradictory visions can emerge in opposite limits of one and the same theory. At least this is the merchandising pep talk. In this quest, people see the need to go to Planck energies which is entirely ridiculous IMO since neither theory says something about this kind of exotism (it seems to me that one should better start looking for unification at distance scales which are at least a factor of 10^{14} higher). To keep this short, I will comment on some remarks made by ``das schones madchen´´ to join the discussion.

**
Another reason why gravity should be quantized: singularity avoidance. The appearance of singularities seems to tell me that we are doing something wrong. However, though this seems a good motivation for me, it is not so completely strict.**

Well it is good that you at least mention that this is not a strict reason at all. Actually, it is far from clear that a theory of QG (whatever this is) is going to avoid singularities. In the same spirit, it is by far not said that nature imposes a dynamical UV cutoff.

** I.e. modifications on a classical level (meaning non-quantized but beyond standard) might do that. **

Well, I know a method which brakes the equivalence principle and does this (you know of any other?). The point I want to make is that one should not worry at all yet about -say- black hole singularities: the energy scales involved are way too high to get any experimental indication about this for the moment. IMO, it is much better to start from an effective unifying theory which is *clearly* open to fasification, that is the only way we can learn something and going down to the Planck scale right away postpones such criterion for at least another century.


**I totally agree with Stingrays argument, just ask: what is the gravitational field of a quark? **

I did not read this argument, but it is for sure not the main question to ask. A better one would be : how much is gravity going to change the EM field of the quark and on what distance scales ? (the compton scale)

** However, also that does not necessarily mean that gravity itself has to be quantized, just that we don't really understand QFT in curved space or its observables. BTW, intersting paper about that**

We do not even have a realistic, rigorous QFT in FLAT space time which is known to satisfy the Wightman axioms AFAIR. In curved spacetime the situation is even far worse...

**
However, what is clear is that gravity at the Planck scale is a key point for our understanding of a unified theory **

No, I disagree ... very interesting gravitationally induced effects already show up at the Compton scale of elementary particles. The Planck scale is not needed at all.


**And therefore pretty addictive - at least for me.
What if instead of quantizing gavity, gravity is the CAUSE for quantization?**

Exactly (but you might want to include EM) ! And if you want to understand that, you better start out from *classical* gravitation and EM and *concrete* test models (as some crazy person half a century ago, I am for example not convinced that we *need* the strong nuclear forces at all as a fundamental building block). Of course, this is not done because you draw then the local realist card (as I do) and that is bad, bad, bad ... There are many who support such ideas but very few who actually *do* it.

By the way, you have a really cute nose

Cheers,

Careful

 

[selfAdjoint]

Careful, what I take from this post is that YOU have a theory, effective at the Compton scale, which "brakes (sp? breaks?) the Equivalence Principle" and in some way unifies EM and GR. I know a couple of theories that claim to do things like that, for example Schifflet's development of http://www.artsci.wustl.edu/~jashiffl/einstein-schrodinger.html" . Without breaking our guidelines could you indicate the basic approach of your theory? If you don't want to air it here, you could PM me.

 

[Careful]

Careful, what I take from this post is that YOU have a theory, effective at the Compton scale, which "brakes (sp? breaks?) the Equivalence Principle" and in some way unifies EM and GR. I know a couple of theories that claim to do things like that, for example Schifflet's development of http://www.artsci.wustl.edu/~jashiffl/einstein-schrodinger.html" . Without breaking our guidelines could you indicate the basic approach of your theory? If you don't want to air it here, you could PM me.

As always you are too enthousiastic If I would have such theory already, you could for sure find it on the arxiv. I did not know about Schifflet but the equivalence breaking mechanism which avoids singularities (this is not a theorem but there is some evidence) can be found in the work of Shojai (not that I put my hand in the fire for this, far from) which is a de Broglie type approach to QG. That ordinary gravitation can drastically modify EM (through the Einstein - Maxwell equations) at the Compton scale of elementary spinning particles (such as the electron) was already mentioned by myself before (actually, this was already known at the end of the sixties...). I remind you that this was simply a reaction against the usual statement that new effects (in comparison to the flat spacetime approximation) are to be found at the Planck scale - which is rubbish.

Cheers,

Careful

 

[hossi]

Hi Careful,

Therefore, QG is the name given to the Platonic desire to know how two contradictory visions can emerge in opposite limits of one and the same theory.

Sounds like a good definition... can we make that 'seemingly' contradictory?

people see the need to go to Planck energies which is entirely ridiculous IMO since neither theory says something about this kind of exotism

That's why one should think about it.

Well it is good that you at least mention that this is not a strict reason at all. Actually, it is far from clear that a theory of QG (whatever this is) is going to avoid singularities. In the same spirit, it is by far not said that nature imposes a dynamical UV cutoff.

Right. But it seems to me that singularity avoidence will lead towards the question of renormalization. At least we know that physcical quantities of objects around us are not usually infinite, so nature knows how to deal with divergences.

The point I want to make is that one should not worry at all yet about -say- black hole singularities: the energy scales involved are way too high to get any experimental indication about this for the moment. IMO, it is much better to start from an effective unifying theory which is *clearly* open to fasification, that is the only way we can learn something and going down to the Planck scale right away postpones such criterion for at least another century.

I think you are right about the importance of effective theories but it is not the only way and it should not be the only way considered.

We do not even have a realistic, rigorous QFT in FLAT space time which is known to satisfy the Wightman axioms AFAIR.

At least it works.

However, what is clear is that gravity at the Planck scale is a key point for our understanding of a unified theory **
No, I disagree ... very interesting gravitationally induced effects already show up at the Compton scale of elementary particles. The Planck scale is not needed at all.

Depends on what you mean with unified theory: you mean the 'effective unifying theory' you mentioned above? Anyway, I admit: I use Planck scale as a synonym for 'the scale at which effects of quantum gravity become important', so you might have a point here.

you better start out from *classical* gravitation and EM and *concrete* test models (as some crazy person half a century ago, I am for example not convinced that we *need* the strong nuclear forces at all as a fundamental building block)

EM does not so really excite me, but yeah, concrete test models, phenomenology, let nature decide, blabla, exactly

On the other hand it would be good to see how such effective models can follow from some fundamental theory.

Concerning my nose: it is pretty red, seems as if I brough a cold from Canada :yuck:

Take Care,

Sabine

 

[Careful]

Hi Hossi,


**Hi Careful,
Sounds like a good definition... can we make that 'seemingly' contradictory? **

No, they are strictly speaking contradictory.


**That's why one should think about it. **

I disagree, it is not the right time to think about this. There is no experimental evidence for physics at those scales at all, and people tend to forget that IF there were something like a *lattice - like* Planck scale cutoff, then a simple electron would involve something like 10^{56} interacting degrees of freedom. That is way beyond anything we are capable off


**Right. But it seems to me that singularity avoidence will lead towards the question of renormalization. **

I think that *physically* renormalization is not even the right kind of question to ask in a satisfying theory with a realistic *particle* notion (singularities or no singularities).


** At least we know that physcical quantities of objects around us are not usually infinite, so nature knows how to deal with divergences. **

Perturbatively in general (so that does not say anything strictly speaking) : example, the weak interactions.

** I think you are right about the importance of effective theories but it is not the only way and it should not be the only way considered. **

I disagree again (I was once also of Planck scale signature); it is actually the only way we can LEARN something.

**
Depends on what you mean with unified theory: you mean the 'effective unifying theory' you mentioned above? Anyway, I admit: I use Planck scale as a synonym for 'the scale at which effects of quantum gravity become important', so you might have a point here. **

This effective theory is the best we can reasonably do; and it should be a priority to first find at least ONE (that is how physics used to work )


** EM does not so really excite me, but yeah, concrete test models, phenomenology, let nature decide, blabla, exactly **

That might be a big mistake, all the interesting effects which are known are of course electromagnetic in origin (since the mass/charge ratio of elementary particles is so damn small in geometric units). However, the gravitational potential/unit mass = EM potential/unit charge at the Compton scale of elementary spinning particles - say -. That is why gravitation CAN influence the EM force (and quite drastically - equal charges strongly attract on these scales ). So, if you want to: gravitation works indirectly through EM. You cannot just focus on one field in physics, you need to take all forces into account (that is one of the mistakes of LQG and all these pure gravitation oriented approaches).

**
On the other hand it would be good to see how such effective models can follow from some fundamental theory. **

But such effective theory IS the fundamental one by lack of any further experimental input !

**
Concerning my nose: it is pretty red, seems as if I brough a cold from Canada :yuck: **

Ah, cure it well

 

[hossi]

Hi Careful,

why the name? I don't get the impression you are really careful - as long as we can't verify or falsify anything, how can you be absolutely sure? That's why I think it's not wise to put all bets on one topic, but to keep the mind open to various approaches. However, it is surely a matter of discussion which one is the 'best' way (and a question of personal taste which one to choose). I understand that you are convinced of your approach -- which is absolutely necessary! -- but coexistence is a huge source for inspiration.

No, they are strictly speaking contradictory.

You are disagreeing with yourself. How strictly can they be contradictory, when you say, there is something to LEARN which might help us understand why General Relativity and Quantum Theory work in some limits? They are contradictory if you try to combine them in a naive way - there are plenty of examples that failed.

I disagree, it is not the right time to think about this. There is no experimental evidence for physics at those scales at all, and people tend to forget that IF there were something like a *lattice - like* Planck scale cutoff, then a simple electron would involve something like 10^{56} interacting degrees of freedom. That is way beyond anything we are capable off

Umpf, that seems to me a very weird argument. If you assume the electron is 'simple' as you say, then it better interacts more simply. However, you are right that it's a big point how matter itself should be described.

I disagree again (I was once also of Planck scale signature); it is actually the only way we can LEARN something.

Depends on what you want to learn, doesn't it. I am not a big fan of string theory, but I find the question what form a so-called theory of everything possibly could take kind if interesting.

This effective theory is the best we can reasonably do; and it should be a priority to first find at least ONE (that is how physics used to work )

I agree with that totally. However, one should not forget that progress in physics very often goes along with progress in Mathematics and vice versa. I think the one can't do without the other. E.g. Mathematics of Yang-Mills-Theories, Einstein and Tensor Calculus, Gell-Mann and Lie-Groups etc, there a more examples.

That might be a big mistake,

I am sure I have made bigger mistakes.

So, if you want to: gravitation works indirectly through EM.

Listen, I have no idea what you mean with that. Is that a result from whatever model or is this what you think it should be or is this just the general statement that on a very fundamental level both actually should be the same thing?

You cannot just focus on one field in physics, you need to take all forces into account (that is one of the mistakes of LQG and all these pure gravitation oriented approaches).

It's a reasonable point to start from. I think they will try to address the other interactions rather sooner than later.

But such effective theory IS the fundamental one by lack of any further experimental input !

That is a pretty philosophical question as long as you don't define what you mean with 'effective'. Maybe we are just talking about the same thing. You are talking about a theory that actually describes everything that can be tested by experiment?

Take Care,

Sabine

 

[Careful]

**Hi Careful,
why the name? I don't get the impression you are really careful - as long as we can't verify or falsify anything, how can you be absolutely sure? **

Oh, but I am very careful. Look, in physics you never can be *absolutely* sure about something but you can make statements at a high confidence level. To be absolutely sure is something which has meaning only in the platonic world of mathematics (most physicists actually find this concept silly). For example, try to tell a physicst that ``you are not absolutely sure that Newtonian mechanics has been falsified´´and you think Newtonian theories should be intensively researched again. Strictly speaking you are correct: there is no no-go theorem against Newtonian mechanics, but I am afraid the physicist will react a bit differently

** That's why I think it's not wise to put all bets on one topic, but to keep the mind open to various approaches. **


If you do not bet your money on one approach, you get nowhere with your results (but your career flourishes at least in QG) ! Progress is only made by hard, consistent work on one ONE topic (actually in mathematics it is ok to be pretty ignorant of other fields).


**
However, it is surely a matter of discussion which one is the 'best' way (and a question of personal taste which one to choose). **

That is not true! The approaches have to satifsy general criteria of good theoretical PHYSICS research (and I am afraid that limits the number of approaches severely )


** I understand that you are convinced of your approach -- which is absolutely necessary! -- **

I am not convinced at all that it will work, but I am pretty sure that it is more or less the only approach which (a) is logical and (b) can produce falsifiable results (and therefore insight) in a reasonable time span.

**but coexistence is a huge source for inspiration.**

I am afraid that we live in times where it distracts (!). What is the gain you get from let's say a seminar on spin foams where the author is explaining the emergence of some nice abstract mathematical structure assuming some axiomatic rules concerning the dynamics without even justifying or giving an example why we should believe this to relate to well tested continuum physics? It is fun when you go to such seminar once, but it gets really annoying when you notice that these things happen already for over five years.


**You are disagreeing with yourself. How strictly can they be contradictory, when you say, there is something to LEARN which might help us understand why General Relativity and Quantum Theory work in some limits? **

No, I do not contradict myself logically speaking (as is easy to deduce).
Any theory of quantum gravity is at least disjoint with either quantum theory or classical gravitation (or either both). The proof that they are incompatible is Bell's theorem.

**
Umpf, that seems to me a very weird argument. If you assume the electron is 'simple' as you say, then it better interacts more simply. **

By ``a simple electron´´ I meant something which is a silly as an electron compared to the scale of the universe (I guessed that was clear from the context).

** Depends on what you want to learn, doesn't it. **

No, it does not. Look, we are not going to learn ANYTHING from something which starts at the Planck scale ; that is a simple matter of counting degrees of freedom and our ability to understand complex systems. Actually, the last 20 years of theoretical physics confirm my thesis.

**I am not a big fan of string theory, but I find the question what form a so-called theory of everything possibly could take kind if interesting.**


That is interesting, but philosophy and way out of reach !


**I agree with that totally. However, one should not forget that progress in physics very often goes along with progress in Mathematics and vice versa. I think the one can't do without the other. E.g. Mathematics of Yang-Mills-Theories, Einstein and Tensor Calculus, Gell-Mann and Lie-Groups etc, there a more examples. **


Agree to a certain level. The big physical discovers actually NEVER started out from the abstraction mathematicians praise, they all started out from elementary coordinate calculations. You can do relativity without modern differential calculus (Weinberg proves that) and you can understand Yang Mills theory without ever having heard about a fibre bundle.

**
Listen, I have no idea what you mean with that. Is that a result from whatever model or is this what you think it should be or is this just the general statement that on a very fundamental level both actually should be the same thing? **

This is an old concrete result about Einstein Maxwell theory.

** It's a reasonable point to start from. I think they will try to address the other interactions rather sooner than later. **

It is not (if you want to I can elaborate on that).

**That is a pretty philosophical question as long as you don't define what you mean with 'effective'.**

By effective I simply mean any kind of theory which starts out from matter configurations involving scales which are possibly far larger than the Planck scale.

Cheers,

Careful

 

[hossi]

Look, in physics you never can be *absolutely* sure about something but you can make statements at a high confidence level.

Yup. It seems to me your confidence level is quite high.

Strictly speaking you are correct: there is no no-go theorem against Newtonian mechanics, but I am afraid the physicist will react a bit differently

Well, it just does not describe nature in certain limits. It is outruled by experiments. Not so with LQG or Stringtheory or whatever. So what would justify to just abandom one of these approaches? Because you say so?

If you do not bet your money on one approach, you get nowhere with your results (but your career flourishes at least in QG) ! Progress is only made by hard, consistent work on one ONE topic (actually in mathematics it is ok to be pretty ignorant of other fields).

I have no idea how you got that opinion. My experience is that your career flourishes the more you specialize to one topic, become the real expert, preferably (but not necessarily) come up with some specific subtopic you can call your own, climb up the connection and community ladder and live happily ever after.

In young fields, there is often some disagreement about the right way to go, if that is what you mean.

Most importantly: you completey neglect the question where the topic is supposed to come from.

However, it is surely a matter of discussion which one is the 'best' way (and a question of personal taste which one to choose). **
That is not true! The approaches have to satifsy general criteria of good theoretical PHYSICS research (and I am afraid that limits the number of approaches severely )

Yes, I agree on that. However, who is going to say what is good and what is bad? And are we in the position to tell right now which QG approach is the best. I don't think so.

I am not convinced at all that it will work, but I am pretty sure that it is more or less the only approach which (a) is logical and (b) can produce falsifiable results (and therefore insight) in a reasonable time span.

If it is the only logical approach, what if it fails?

What is the gain you get from let's say a seminar on spin foams ...

Inspiration? Coffee and Cookies? Improve my English? Meet new people? Hear new opinions? Actually, I don't sleep well at night, but nothing is as soothing as a really bad seminar speaker.

It is fun when you go to such seminar once, but it gets really annoying when you notice that these things happen already for over five years.

So 5 years is your time-out?

No, I do not contradict myself logically speaking (as is easy to deduce).
Any theory of quantum gravity is at least disjoint with either quantum theory or classical gravitation (or either both). The proof that they are incompatible is Bell's theorem.

So there is no QG?

** Depends on what you want to learn, doesn't it. **
No, it does not. Look, we are not going to learn ANYTHING from something which starts at the Planck scale ; that is a simple matter of counting degrees of freedom and our ability to understand complex systems. Actually, the last 20 years of theoretical physics confirm my thesis.

How do you know how to count the degrees of freedom without knowing what is going on? I do agree with you that the Planck Scale might not be a good point to start from since we just don't know enough to make that work. It's like poking around in the dark: just unlikely to find something.

That is interesting, but philosophy and way out of reach !

I believe philosophy to be a branch each society needs to have. However, you might state that this is not physics, which then leads to the question where to draw the line between maths, physics and philosophy. That is a matter one could easily write a 400 pages book about and I leave that to someone who speaks proper English.

** It's a reasonable point to start from. I think they will try to address the other interactions rather sooner than later. **
It is not (if you want to I can elaborate on that).

Yeah, I would really like to hear your point of view.

Take Care,

Sabine

 

[Careful]

**
Well, it just does not describe nature in certain limits. It is outruled by experiments. **

That is not true AFAIK, eather theories are strictly speaking not ruled out by experiment, although the constructions involved are sometimes so contrived that they become highly implausible (and that is exactly what I wanted to argue ).

**
Not so with LQG or Stringtheory or whatever.**

Again, these are not theories satisfying the criteria of good physics as you should know. A good physical theory (a) is mathematically well defined (b) makes predictions consistent with current observation (c) makes new fasifiable predictions (d) has a very limited number of free parameters... I hope I do not have to specify any further why the above candidates are ruled out by these four laws.


**) come up with some specific subtopic you can call your own, climb up the connection and community ladder and live happily ever after.**

I was talking about QG in particular.

**In young fields, there is often some disagreement about the right way to go, if that is what you mean. ¨¨


You should remember your history. The quest for QG is eighty years old (!). String theory is over thirty, ``the canonical approach´´ is over twenty plusminus the same as causal sets, dynamical triangulations are over thirty again... NEVER in history has there such been an intensive search (if you calculate the number of researchers times the time invested then this is the equivalent of more than a few centuries ...), and we really did not get far at all... So, the field is not young by any means.

**
Yes, I agree on that. However, who is going to say what is good and what is bad? And are we in the position to tell right now which QG approach is the best. I don't think so. **

Any approach which gives real predictions (not some handwaving stuff people shake out their hat) will be automatically numero uno.

**If it is the only logical approach, what if it fails? **

Then we have learned that something else is needed (which I doubt very much). In the best case: we can also identify a reasonable factor where it goes wrong and then try to fix it. That is how science progressed since eternity...


**
Inspiration? Coffee and Cookies? Improve my English? Meet new people? Hear new opinions? Actually, I don't sleep well at night, but nothing is as soothing as a really bad seminar speaker.**

I see you will make your way to the top :rofl:


**So 5 years is your time-out? **

5 to 10 years is a reasonable time for a theory to catch up with reality no ? I do not even ask to make a new prediction.


** So there is no QG? **

HUH ? Sorry, but I do not understand what is so difficult to comprehend in what I say here. I made a formulation AA of QG and you wanted to add the word seemingly contradictory (to both theories emerging in different limits); I disagree because of Bell's theorem but this does not imply that AA does not exist !

**
How do you know how to count the degrees of freedom without knowing what is going on? **

I counted ONE degree of freedom per Planck length in each direction; so this is a reasonable LOWER bound. If you want to increase the number of degrees of freedom, fine, that will only add strength to my argument

** I do agree with you that the Planck Scale might not be a good point to start from since we just don't know enough to make that work. It's like poking around in the dark: just unlikely to find something. **

Exactly, so bye, bye Planck scale approaches...

**I believe philosophy to be a branch each society needs to have. However, you might state that this is not physics, which then leads to the question where to draw the line between maths, physics and philosophy. That is a matter one could easily write a 400 pages book about and I leave that to someone who speaks proper English.**

Exactly my point; you are interested in books of 400 pages how the quest to a certain ideal should be perceived

**
Yeah, I would really like to hear your point of view.
**

I will come back to this later, have to go now...

 

[hossi]

Sabine and the chocolate factory

That is not true AFAIK, eather theories are strictly speaking not ruled out by experiment, although the constructions involved are sometimes so contrived that they become highly implausible (and that is exactly what I wanted to argue ).

Okay, I was referring to 'pure' Newton (no eather or something funny).

Not so with LQG or Stringtheory or whatever.**
Again, these are not theories satisfying the criteria of good physics as you should know. A good physical theory (a) is mathematically well defined (b) makes predictions consistent with current observation (c) makes new fasifiable predictions (d) has a very limited number of free parameters... I hope I do not have to specify any further why the above candidates are ruled out by these four laws.

I agree with you on that. Actually, I find the situation kind of funny because I share your opinion. Things are going seriously wrong and something has to be done about it.

However, I don't think going to the complete other extreme and abandom theories that do not fulfill your points is a wise decision either. But I do think that resources (meaning money and people) should be distributed in accordance to your points a) - c), which definately is not the case right now (point d) is too involved to start with).

But apparently you want to hear points on which to disagree rather than to agree. So, the reason why I plead for tolerance is related to Nestle. I can't stand Nestle-tasting stuff, no matter what. Chocolate, icecream, yoghurt, whatever. Unfortunately, there seem to be many people who like it (or who just have never tried anything else.) As long as I have alternatives, good. But I am constantly afraid, at some point Nestle will eat up all other smaller companies who don't sell too well, close them and I will have to live with Nestle-taste the rest of my life. :yuck:

To come back to the point, monopols do endanger quality and are on the long run an obstacle to progress. Not to mention that it makes things plain boring.

You should remember your history.

I wasn't aware I had one...

The quest for QG is eighty years old (!). String theory is over thirty, ``the canonical approach´´ is over twenty plusminus the same as causal sets, dynamical triangulations are over thirty again... NEVER in history has there such been an intensive search (if you calculate the number of researchers times the time invested then this is the equivalent of more than a few centuries ...), and we really did not get far at all... So, the field is not young by any means.

You are neglecting an important point: the explosion of work to one topic in a very short time makes it almost impossible to sort out things. There is a severe lack of people who a) can judge on what makes sense b) have and overview and c) whose opinion is accepted by the community. There are developments which need time that you can not push with just more people because they will loose contact (maybe there can be some technical improvement made about that, e.g. the arxiv could have a comment option or so, but that won't completely solve the problem).

That is a development which is new in science and a problem that has never been there before.

Any approach which gives real predictions (not some handwaving stuff people shake out their hat) will be automatically numero uno.

Sure. That is not the question. The question is what to do as long as this approach is not available.

**If it is the only logical approach, what if it fails? **
Then we have learned that something else is needed (which I doubt very much).

That does not make sense. Something else would mean it was not the only logical approach to begin with.

[...] so bye, bye Planck scale approaches...

we can also identify a reasonable factor where it goes wrong and then try to fix it. That is how science progressed since eternity...

Maybe. But you are making a mistake that most politicians also make. You assume that a way is right because it was always the right way before.

Let me give you an example because I am in a chatty mood right now Let us suppose evolution a la Darwin is right (to avoid a discussion about that). Why are there no animals on wheels? Aren't wheels a great thing which contributes a huge lot to the 'fitness' of human beings ('fitness' in Darwins sense only I am afraid).

Two reasons: 1) A wheel only works if it has all components together in the right way, its not suffiecient just to be round, it needs an axis and some way to navigate etc. It is fairly complex and thus unlikely (though not impossible) to be produced by evolution.

2) A wheel works best on proper ground. You can produce that most easily by driving the same way lots of times. Or you can go high tech and pave it.

Toghether, wheels are not likely to appear in evolution. But we DO have wheels and we use them to keep our infrastructure on a high level. How did that come along?

Nature took the sneaky way: there were these humans with brains so large that at some point they were not busy any more the whole day hunting and planting and so on. So, some of them would sit around for some time doing weird stuff, while the other ones did the real thing like feeding the kids or protecting the women from bears.

I guess, some of the weird ones got probably eaten up by the same bears while they were sitting around. Or they starved or whatever. Some of the real guys must have taken pity on them, proteced them and gave them the last deer. Who would have thought at that time, the weird ones could come up with something as useful as a wheel? But eventually they did! However, even when they had invented a wheel, they might not have known HOW useful it was until they figured out how to level their paths etc.

So, why was I telling that?

Oh, yes: Though there might be a straightforward road to progress that road might be very hard to go and take a very long time. Progress can also be made by jumping off that road and finding shortcuts that were not foreseeable. Bottomline: feed the weird ones, they could turn out to be useful.

HUH ? Sorry, but I do not understand what is so difficult to comprehend in what I say here. I made a formulation AA of QG and you wanted to add the word seemingly contradictory (to both theories emerging in different limits); I disagree because of Bell's theorem but this does not imply that AA does not exist !

Yeah, I guess that discussion is not going anywhere. I mean they are seemingly contradictory because the existence of QG means that they both actually can be formulated toghether in a selfconsistent way. So, Bells theorem makes them appear contradictory because we have not understood how the limits of GR and QM do arise and how they are related to the full thing.

I counted ONE degree of freedom per Planck length in each direction; so this is a reasonable LOWER bound.

Is it...? Haven't you already made an assumption that you can not be sure of?

If you want to increase the number of degrees of freedom, fine, that will only add strength to my argument

I surely would not want to INcrease them.

Anyway. Gotta work. Nice talking to you

Sabine

 

[Careful]

Jee, you are chatty today :rofl: I will be brief since it is bed time over here and I shall only react upon the points where I disagree just because I am a grumpy person.

(a) A semantic issue about the incompatibility of QM and GR: I guess we both agree upon what it means but we disagree upon the formulation. So, let's quit that one, it is getting boring.

(b) How to reach QG? My point of view is very clear: we have NO experimental guideline whatsoever (at least Einstein did have some indications for GR) so we are entirely lost. The good way to proceed in this case then is to radically sacrifice one of both theories and build upon the other one, i.e. see how far one theory ``deviates´´ from the other.
(1) LQG, STRT and CDT builds upon QM and it is far from sure that they get out GR in the classical limit.
(2) Local realism starts from GR and has to get out microscopic physics (note: this is NOT necessarily equal to QM).

Approach (1) has a long very active history and according to me, a key problem on the road there is the measurement (or micro - macro) problem in QM with the bourlesque name: Schroedinger's cat. This problem is especially important since I want to know how the macroworld, with its local realism can be understood to emerge from a microscopic world with an operational ontology. IMO, this problem is unsolvable within contemporary QM and little or no progress has been made in this field.
Penrose OR proposal is one of the more intelligent proposals I have heard concerning this problem. So (1) has a really bad record if you remember my list of four laws.

Now (2) is not experimentally falsified and it is really a mystery for me why people have not put more effort into this:
(a) there are no ontological problems whatsoever (so the approach is self consistent)
(b) spacetime locality and covariance are the MOST simplifying principles to construct a theory
(c) There EXIST already results which reasonably indicate that local realism (LR) can reproduce quantum mechanical results (think about stochastic electrodynamics and the Barut self field approach)
...


The reason why (2) is unpopular I guess is because it sacrifies QM (which is outcasting GR). Moreover, the deliberate misinformation about Bell test results is not really contributing to an honest evaluation (I have still some unfinished business on the QM forum). So, the best you can reasonably do is propose concrete test models and TEST them That is actually Einsteins methodology which has proven to be highly succesful.

So, I am not contradicting myself here: the first *logical* step is to choose radically for one theory and option (2) is leading much faster to experimental falsification. Actually, this is the big *advantage* of (2) : we have pleanty of experimental results to test against: check out any book on quantum chemistry. Then, when it would turn out that (2) is reasonably not possible (which would require testing lots of continuum models for elementary particles and so on), *then* we might think about CAREFULLY abandonning spacetime locality (and not that wild as happens in QM). Next, if we would find such deviation which is in concordance with all experimental results, THEN we could try to derive this theory from basic new physical principles. This seems to me the road to succes.


** Haven't you already made an assumption that you can not be sure of?
I surely would not want to INcrease them. **

Sure, but now you should think of other models available on the market and do the (rough) counting there (you shall see there is no substantial deviation). You must admit that it is a fair way to get a rough estimate (and that is all I wanted to provide).

Your Nestle and wheel analogies can be easily answered but that will have to wait. It is nice talking to you too

Cheers,

Careful

 

[hossi]

Hey Careful,

(c) There EXIST already results which reasonably indicate that local realism (LR) can reproduce quantum mechanical results (think about stochastic electrodynamics and the Barut self field approach)
...
The reason why (2) is unpopular I guess is because it sacrifies QM (which is outcasting GR). Moreover, the deliberate misinformation about Bell test results is not really contributing to an honest evaluation (I have still some unfinished business on the QM forum).

I have no proplem with sacrifying QM - I don't particularly like it.

Could you give me some references on the above (preferably on the arxiv)?

Thanx, sleep well,

S.

 

[Careful]

Hey Careful,
I have no proplem with sacrifying QM - I don't particularly like it.
Could you give me some references on the above (preferably on the arxiv)?
Thanx, sleep well,
S.

Tomorrow: most of them are Phys Rev though (but I assume you have easy acces to that too). Let me add the obvious which is that altough these approaches make correct predictions in some cases, they are also known to deviate *slightly* (which is actually hopeful) from experiment in other circumstances (this despite the fact that entanglement is *entirely* given up).

:zzz: :zzz: :zzz:

Here are some SEMICLASSICAL approaches (it is not what I have ultimatly in mind, but it is a good start)

Phys rev A, 55, 3879 (1997) Fourth-order interference in the Wigner representation for parametric down - conversion experiments

Phys rev A, 39 (6), 2796, Quantum electrodynamics based on self-fields, without second quantization
Phys rev A, 38 (9), 4405, idem


European Physics Journal D, 1, 317-327 Semiclassical theory of Compton and photoelectric effects

 

[Maaneli]

To quantize or not?

In fact, one does not need to quantize the spacetime manifold to derive the Einstein-Hilbert action. One need first assume a flat Lorentzian manifold, and perform one-loop quantum field effects to generate the effective dynamics of standard GR. See the following works of Visser, Liberati, and Barcelo, who extend the ideas of Sakharov that "Einstein gravity is an emergent low-energy long-distance phenomenon that is insensitive to the details of the high-energy short-distance physics."


http://arxiv.org/abs/gr-qc/0204062

http://arxiv.org/abs/gr-qc/0111111

http://arxiv.org/abs/gr-qc/0106002

Cheers,
Maaneli

 

[vanesch]

Approach (1) has a long very active history and according to me, a key problem on the road there is the measurement (or micro - macro) problem in QM with the bourlesque name: Schroedinger's cat. This problem is especially important since I want to know how the macroworld, with its local realism can be understood to emerge from a microscopic world with an operational ontology. IMO, this problem is unsolvable within contemporary QM and little or no progress has been made in this field.

Now, *that* I don't agree with! An MWI view (with a few additional postulates about the link between the ontology of the world and the subjective perception of it) CAN do it. You may not like it (you call it "believing in ghosts"), you may be convinced that it is not the right approach, but you cannot say that there is no solution to the problem of Schroedinger's cat without (non-local) collapse.

You should acknowledge this, as much as it should indeed be acknowledged that LR has not been outruled by experimental results.

 

[Careful]

Now, *that* I don't agree with! An MWI view (with a few additional postulates about the link between the ontology of the world and the subjective perception of it) CAN do it. You may not like it (you call it "believing in ghosts"), you may be convinced that it is not the right approach, but you cannot say that there is no solution to the problem of Schroedinger's cat without (non-local) collapse.
You should acknowledge this, as much as it should indeed be acknowledged that LR has not been outruled by experimental results.

Hi vanesch,

Well, the MWI approach is (a) incomplete (give me a theory of consciousness and state projection; when does it take place etc...) (b) highly uneconomic (parallel universes, conscious beings, zombies and all that). Moreover, macrorealism means a bit more than ``why is the cat perceived alive or dead´´: for example one must also show that the conscious observations (in different universes) must satisfy more or less locally causal laws of motion (no entanglement). Anyway, more information is to be found in the paper of A. Leggett which I referred you to by PM a while ago.

Cheers,

Careful

 

[vanesch]

Hi vanesch,
Well, the MWI approach is (a) incomplete (give me a theory of consciousness and state projection; when does it take place etc...) (b) highly uneconomic (parallel universes, conscious beings, zombies and all that). Moreover, macrorealism means a bit more than ``why is the cat perceived alive or dead´´: for example one must also show that the conscious observations (in different universes) must satisfy more or less locally causal laws of motion (no entanglement).

I think you still didn't get exactly the meaning of MWI! But we're not going to have this discussion here again, right...

Anyway, more information is to be found in the paper of A. Leggett which I referred you to by PM a while ago.

Yes, I tried to find that with the info you PMed me, but couldn't. Any more precise reference of it ?

 

[Careful]

**I think you still didn't get exactly the meaning of MWI! **

Ah, you might have been confused by my use of ``state reduction´´ (which is my personal terminology for the assignment of consciousness to a particular physical state).

** But we're not going to have this discussion here again, right... **

No need to, it is impossible to forget such story which had the same impact on me as Haensel und Gretchen of the Grimm brothers

**
Yes, I tried to find that with the info you PMed me, but couldn't. Any more precise reference of it ?**

Testing the limits of quantum mechanics : motivation, state of play, prospects. J. Phys, condens matter 14, R415-R451

You can get it for free at the homepage of Leggett. You will notice that he is less friendly to the relative state interpretation than I am : ``a meaningless collage of words´´ to quote the good man.

Cheers,

Careful

 

[selfAdjoint]

If this is going to turn into yet another discussion of MWI, doesn't it belong on the quantum physics forum?

 

[Chronos]

So it would appear, Vanesch. Would a lamp appear to be too Dionysian?