Is Entropic Gravity the Future of Physics?

[Dmitry67]

Question

How holographic principle is defined if we don't use null surfaces of the region - but regular ones, timelike? So normal boundaries we use (a box, a sphere...)

The problem I see: when 'region' is spacially extended, the position of the boundary (and the very notion of the region) becomes frame-dependent (in SR)

It is even worse in GR and very curved spaces. Talking about the Dark Energy and expansion (this is why I am asking this question) I don't see how 'region' can be bigger than a Hubble volume.

P.S. I assume that Dark Energy is enropic force associated with entropy of geometry of our spacetime?
If our space is so surprisingly flat, then it should tend to transform into more chaotic state...

 

[Fra]

I've been low on time lately, but IMHO, beeing am a advocate of a view that includes kind of "entropic forces" (as another expression of rational action which is the idea that the action of any systems basically follows "random" motion in a evolving state space) but I still see view Verlinde's reasoning so far so still lack the deeper pictures.

I think it's possible withfor alternative developments of the initiated route and the ultimate version of the arugments are still not on the table.

It seems all different people try to interpret or construct a connection to their advantage. I am so far not so much liking smolins connection. I would rather want to see a smolin connection to verlind á la R. Unger, rather than á la LQG.

/Fredrik

 

[MTd2]

Hey Marcus! Well, I was not joking. But saying geometry should not be quantized is not my idea, but Verlide's! That's what I understood. He could come out saying whatever he wanted in NYT, so that layman could understand, except for things like "X" does/doesn't exist. It`s like saying "cancer is not a unique desease" or "HIV causes AIDS". This is something that can be understood by anyone...

 

[Fra]

In any case, if you seriously credit the idea that one should not quantize geometry, then you must be willing to make predictions. Perhaps you would like to predict a decline in QG publication/cites. And an increase in research where one looks at "how quantum fields affect classical geometry".

I apologize if I miss the point as I haven't had time to read all the papers discussed but I seek at least two ways to interpret "one should not quantize geometry or gravity".

research where one looks at "how quantum fields affect classical geometry".

- One possibility is that "gravity" is simply a classical thing, that exists only (in a sensible way) in a classical limit and in the true quantum domain there is no gravity? In the sense that statistical laws doesn't exist at microlevel, and gravity is only emergent at classical and macroscopic level?

- The other possibility (to which I adhere) is that it's perhaps just to say that it's possibly the wrong approach to the problem of QG to start to look for a regular "quantization procedure" of a classical thing or field. But this is also a critique agains the entire framework of how QM is usually introduce - ie from a classical model, then applying a "quantization" trick. Instead, perhaps a proper reconstruction of a measurement theory, that does NOT start from a classical counterpart, will show that gravity is emergent from the "new quantum framework" itself, as a kind of entropic force. But in this view, IMO at least, the same must apply to ALL forces. ALL forces should be "entropic", the distinction between the various forces would then probably be a matter of classifying components of a generic "entropic force".

Note that even Ariel Caticha, who is working on inference models (a sort of version of entropic reasoing, but still different from Verlinde!) also have said that if he is right the entire approch to the "quantum gravity" problem are likely to need new grip.

This latter view is what I think, and it's why I think "quantizing gravity" in the regular sense is doubtful. But that's not to say that gravity is just classical, it rather means (to me at least) something different.

/Fredrik

 

[marcus]

David Berenstein is a string theorist who is still, AFAIK, pursuing string theory. He hasn't turned to last minute career-saving gimmicks of the "anything to get me out of this rut!" variety. Here's a 13 July 2010 blog post of his.

======quote======
Dirac Sea Shore — Gravity does not exist?

I happened onto an article on the New York Times abut Erik Verlinde’s take on gravity as an Entropic force. The article was written by Dennis Overbye who most of the time does a good job of covering high energy physics. Erik’s work dates from earlier this year and can be found here. To tell the truth, I don’t understand what he’s trying to say in that paper and to me it feels like it’s almost certainly wrong.

However, I don’t want to discuss that paper. What I want to discuss is the following provocative quote

“We’ve known for a long time gravity doesn’t exist,” Dr. Verlinde said, “It’s time to yell it.”

I don’t believe this is taken out of context, so we should take it at face value. The statement is obviously wrong, so it sounds like ultra-post-modern pap and makes all physicists working on the subject of quantum gravity look like crazy mad men. I’m sure this sells newspapers, but that is not the point.

When asked for a sound byte can’t people at least say something that is correct and not just provocative?

The proper way to write that statement is that “Gravity is not really a fundamental force “, which is more correct and does not deny gravity its proper place as something that has been observed in nature, however it is less catchy. If we apply the same criteria as used in the above construction, all of the following statements are also correct:

Hydrodynamics does not exist (it only happens for collections of atoms, but not for individual ones)
Space and time do not exist (often used when talking about quantum gravity being emergent from somewhere else)
All emergent phenomena do not exist (they are not fundamental after all).
I do not exist (I’m an emergent phenomenon)...
===endquote===
http://diracseashore.wordpress.com/2010/07/13/gravity-does-not-exist/
Incidentally, Berenstein may be looking for his own way out of the woods. See for example http://arxiv.org/abs/1001.4509

 

[MTd2]

Marcus, Erik indeed thinks gravity does not exist. Check this radio interview out:

http://www.thetakeaway.org/2010/jul/14/argument-against-gravity/

The way I think out of this is imagining a nonquantized space time where it has a texture of rubber, where it bends according to entropy, not matter. And this rubber, by itself, is made of nothing, so it is like a space time without any self interaction. It is a perfect void.

 

[my_wan]

Marcus, Erik indeed thinks gravity does not exist. Check this radio interview out:

http://www.thetakeaway.org/2010/jul/14/argument-against-gravity/

The way I think out of this is imagining a nonquantized space time where it has a texture of rubber, where it bends according to entropy, not matter. And this rubber, by itself, is made of nothing, so it is like a space time without any self interaction. It is a perfect void.

Gravity does not "exist" is a major distortion of what Verlinde actually said. It was the interviewer who kept representing it that way without correction, whereas Verlinde was merely attempting to convey that gravity wasn't a fundamental force, i.e., it was derivable.

 

[my_wan]

This question of the existential nature of gravity has taken several forms. Started out with Newton, in which it was an existential force. The Einstein represented it as a force free geometry, which could be described as non-existent wrt force. What Verlinde is actually claiming as non-existent is gravity as a 'fundamental' property. It's on par with claiming temperature is non-existent on the grounds it's derivable from mechanics, but spacetime doesn't specify a consistent mechanics to embed the thermodynamics.

Non-existence has taken on different meanings in different context. It's basically semantics.

 

[MTd2]

Yes, right, that's what we understand from his paper. But there is more to that when he says there is no gravity. For example, suppose we have an action for gravity and try to quantize that. But whatever method we have or theory, quantizing gravity does not make sense because there is no microstates for pure gravity, but just microstates for fields on space time. So, there is no graviton (consequently no string theory), no loop, no spin foam that leads to gravity.

 

[Fra]

quantizing gravity does not make sense because there is no microstates for pure gravity

This phrase makes good sense to me - there is no such thing as "pure gravity" in a measurement theory, since all measurements requires a non-trivial complexity with observers.


/Fredrik

 

[marcus]

So, there is no graviton (consequently no string theory), no loop, no spin foam that leads to gravity.

Good! You have effectively made a prediction. It might be stated as: nobody (say in the next 3 years) will come up with a version of LQG (including spin foam models) with the right largescale limit.

I think your position predicts this because, according to your view, that cannot happen. It would be a version of LQG that leads to gravity---leads to the right classical behavior, which is presumably GR or something very close to it. According to you, that line of development must inevitably slow down and come to a halt.

We can see who is right, over the course of some arbitrary timeframe like 3 years (or less if you prefer). I have different views and expectations from yours. I see indications that people are rapidly nearing the point of demonstrating the correct limit behavior. This keeps being verified in more special cases. I could be wrong of course, but my impression is that the positive results are coming more and more frequently. It is a time of rapid progress in the program, not a time of "slowing to a halt".

Certainly matter must be included in the dynamics. This is in progress. So far rudimentary matter can already be included in the kinematics, and also is already included dynamically in LQC models. But I don't know of it being included in the dynamics of the full LQG theory however. Maybe someone can tell us what the situation is in that regard.

Your argument "no graviton therefore no string theory" makes a certain amount of sense. But it does not carry over to LQG.

The graviton is not natural to LQG, one has to work hard to artificially constrain the states to be flat and only then, after a struggle, one gets something that looks like a graviton propagator. LQG is not about a fundamental force, it is a quantum theory of the disturbances of geometry. The characteristic observables are of geometric measurements like area and volume.

It is very natural that the micro DoF that determine area and volume expectations should have entropy and temperature. This is just the picture that Verlinde was talking about when he referred to the heat bath and the entropic force along a polymer chain. Micro DoF able to assemble themselves in various configurations. Spin network dynamics, and at large scale, spin network thermodynamics. Verlinde just didn't happen to mention that possible realization of his idea.

The notion that "no graviton" implies no LQG seems like an odd conclusion.

 

[MTd2]

Hmm. If LQG is not about gravity, what is the "G" doing there? '

 

[marcus]

Like the old saying goes, "gravity=geometry". LQG is obviously about gravity, because it is about geometry (which GR teaches us is the correct way to look at gravity).
QG has come to have the double meaning of quantum geometry/gravity.
You know all this, no need to say it really. No need to kid around either: that's what the G in LQG stands for and you know it.

Personally I don't think flat space exists (it is only more or less well-approximated by dynamic curved geometries) and I don't think any mathematical theory deserves to be called QG unless there is some mathematical entity in the formalism which represents the geometry of the universe.

Or at least the underlying degrees of freedom from which the geometry of the universe emerges.

So field theories, like conventional QFT, which are based on classical flat geometry, cannot be right. I assume that the quantum field theory of all matter fields will eventually be carried over and defined on a quantumdynamic geometry.

Various ways have been proposed to do this. What comes to mind are proposals which involve labeling the nodes and links of spinnetworks with extra matter field labels.

 

[MTd2]

] (which GR teaches us is the correct way to look at gravity).

Almost the correct way because there is no gravity as a fundamental force. There is no gravitational field. Gravity does not have a Lagrangian nor Hamiltonian just like there isn`t any of these for a continuous flux of 1 atom of the oxygen inside H2O.

The nature of things are too different to name the entropic attraction as gravity when you are looking at very tiny levels. It is better then to call LQG as Loop Quantum Geometry, no gravity. It is like call an isolated atom of hydrogen inside H2O as water.

 

[marcus]

... It is better then to call LQG as Loop Quantum Geometry,..

Indeed I often do think of LQG in exactly that way. And for me, that has always meant the same as Loop Quantum Gravity because, as members of the community often point out, gravity IS geometry.

Gen Rel teaches us that the gravitational field is not a field of force or a field of "gravitons". The gravitational field is simply, and nothing else but, the geometry.

Typically (according to vintage 1915 Gen Rel) the geometry is described by an equivalence class of metrics. The essential web of geometric relationships among events that remains after one throws away every specific spacetime manifold and all systems of coordinates.
This is what Gen Rel has been trying to tell us for almost 100 years. Gravity is a responsive geometry on top of which other fields are to be defined. Except in an approximate sense it should not be treated as a force operating in flat geometry (à la "gravitons" of string.)

 

[atyy]

Almost the correct way because there is no gravity as a fundamental force. There is no gravitational field. Gravity does not have a Lagrangian nor Hamiltonian just like there isn`t any of these for a continuous flux of 1 atom of the oxygen inside H2O.

The nature of things are too different to name the entropic attraction as gravity when you are looking at very tiny levels. It is better then to call LQG as Loop Quantum Geometry, no gravity. It is like call an isolated atom of hydrogen inside H2O as water.

In Verlinde's view, gravity=geometry. Hence if gravity is not fundamental, neither is geometry. Both gravity and geometry are emergent.

http://arxiv.org/abs/1001.0785 "Of course, Einstein's geometric description of gravity is beautiful, and in a certain way compelling. Geometry appeals to the visual part of our minds, and is amazingly powerful in summarizing many aspects of a physical problem. Presumably this explains why we, as a community, have been so reluctant to give up the geometric formulation of gravity as being fundamental. But it is inevitable we do so. If gravity is emergent, so is space time geometry. Einstein tied these two concepts together, and both have to be given up if we want to understand one or the other at a more fundamental level."

 

[MTd2]

Well, if that`s the case, we need some kind of spin foam formulation. Spin foams do not have a geometric description of gravity.

 

[marcus]

Well, if that`s the case, we need some kind of spin foam formulation. Spin foams do not have a geometric description of gravity.

I think I understand, and I agree.

Maybe there is a semantic confusion surrounding what Atyy says. The LQG focus is on finding the microscopic DoF from which conventional geometry emerges.

Conventional "diffy manifold" geometry is, indeed considered NOT fundamental.
One looks for something deeper from which the traditional manifold geometry emerges.
And therefore also must gravity emerge.

Spin networks are not conventional geometry (they are combinatorial objects).
Spin foams are not conventional geometry (they are the diagrams describing evolution of spin networks)

OK! So I see what you are saying!

Atyy puts forward the view that neither conventional manifold geometry nor the "force" picture of gravity is fundamental. One response is to say well therefore what we need is LQG with networks and foams, because those things emerge from the combinatorial objects (with which we can moreover do computations.)

 

[MTd2]

Marcus, I have a little dirty homework for you. Tell me the list of finite list of finite subgroups of SU(2). Next, think of surfing and foams.

 

[marcus]

Marcus, I have a little dirty homework for you. Tell me the list of finite list of finite subgroups of SU(2). Next, think of surfing and foams.

The irreducible representations correspond to half-integers...but finite subgroups is something different.
I can't tell you the classification. I would have to walk onto campus tomorrow and look it up at the math library.
I think the classification of the finite subgroups of SU(2) may be available in this book by a former teacher of mine.
Joseph A. Wolf, Spaces of Constant Curvature (1967 McGraw-Hill).

You are younger than I am. Please tell me your idea and don't make me work for it. Probably Heaven will reward you. And in any case I would not succeed in guessing your idea even if I did go look up the SU(2) subgroups.

===================
Wait! I used google and found something more recent about the finite subgroups of SU(2). Maybe everybody knew this except me. It was over my head. Too technical. But seemed quite interesting. The paper is by Wulf Rossman.
http://arxiv.org/pdf/math/0307121

It refers to a discovery by John McKay and also to some followup work by Bertram Kostant (who was at that Banff Lisi workshop)
Here is from the introduction:

"In 1980 McKay announced his astounding discovery that the finite subgroups of
SU(2) are in natural 1-1 correspondence with the extended Coxeter-Dynkin graphs
of type ADE in the following way..."
====================

I have to pass on your homework. But please tell me what you have in mind.

 

[MTd2]

http://www.pnas.org/content/81/16/5275.full.pdf

One of them is E8. This looks like the one with the highest entropy out of all finite subgroups, due its more homogenous distribuition (basically a superposition of 2 600 cell polytope). So, we could we expect that one likely spins to organize around some kind of E8 quantum ressonant state, somewhat similar to phonoms. Make the system colder and colder and we would get E8 broken, because other levels of ressonance would dominate, like the particles of the SM. Well, this is how I think we could get particles from a spin foam...

 

[Fra]

Well, if that`s the case, we need some kind of spin foam formulation. Spin foams do not have a geometric description of gravity.

Or some information theoretic style formulation; which after all, is the natural home of entropic type inferences? But this direction of the quest essentlally conincides with a better understnding of the foundations of QM.

As far as LQG goes, what I seek is a first principle motivation for the spin networks, that does not use backward-arguments from reforulated GR. With some imagination that might be possible, if you instead consider "action networks" that live in more abstract state spaces, but each time I've tried to read up details on that at least from rovelli it's clear to me that is *not* how He sees it.

/Fredrik

 

[my_wan]

Yes, right, that's what we understand from his paper. But there is more to that when he says there is no gravity. For example, suppose we have an action for gravity and try to quantize that. But whatever method we have or theory, quantizing gravity does not make sense because there is no microstates for pure gravity, but just microstates for fields on space time. So, there is no graviton (consequently no string theory), no loop, no spin foam that leads to gravity.

Funny thing is that Verlinde stated in that paper that it was consistent with string theory, as far as he knew. Makes you wonder what entropy is without microstates to...

 

[atyy]

Funny thing is that Verlinde stated in that paper that it was consistent with string theory, as far as he knew. Makes you wonder what entropy is without microstates to...

The graviton is not fundamental in string theory.

 

[MTd2]

Actually, gravity is the first consistency check for string theory (besides having no ghosts):

http://www.superstringtheory.com/basics/basic5a.html

 

[rod_worth]

Here's my two-cents worth of emotional analogy: String Theory is the modern equivalent of epicycles; modify it until it works.

 

[humanino]

CDT---an Entropic Theory of Quantum Gravity
J. Ambjorn, A. Goerlich, J.Jurkiewicz, R. Loll
(Submitted on 15 Jul 2010)

In these lectures we describe how a theory of quantum gravity may be constructed in terms of a lattice formulation based on so-called causal dynamical triangulations (CDT). We discuss how the continuum limit can be obtained and how to define and measure diffeomorphism-invariant correlators. In four dimensions, which has our main interest, the lattice theory has an infrared limit which can be identified with de Sitter spacetime. We explain why this infrared property of the quantum spacetime is nontrivial and due to "entropic" effects encoded in the nonperturbative path integral measure. This makes the appearance of the de Sitter universe an example of true emergence of classicality from microscopic quantum laws. We also discuss nontrivial aspects of the UV behaviour, and show how to investigate quantum fluctuations around the emergent background geometry. Finally, we consider the connection to the asymptotic safety scenario, and derive from it a new, conjectured scaling relation in CDT quantum gravity.

 

[Haelfix]

That abstract and paper is a good example of advertisement in theoretical physics. Its basically the standard CDT story to the T (albeit written well, its one of the better review papers for CDT on the market).

The authors pay lipservice to just about every other approach in the field, and only write 2 handwavey sentences about Verlinde's ideas (which admittedly is the very definition of handwaving in theoretical physics).

Its designed as a 'cite me' paper, most likely to raise some needed funding for better computers or somesuch.

 

[Orbb]

A brief comment about the direct quantization of gravity: I think it does make sense to quantize gravity even perturbatively as long as you restrict yourself to a limited energy regime. It works perfectly fine for computing quantum corrections to classical gravity as an effective field theory. Jacobson was pointing out that quantizing gravity makes no sense in the thermodynamic picture, but I think he has refrained from this statement. It is similar to condensed matter systems: Phonons are quantized sound waves. They are collective excitations and an emergent phenomenon. Still, they exhibit quantum properties. Compare this also to Fermi theory: it doesn't use the 'fundamental' degrees of freedom and is nonrenormalizable, but still it's convenient for some calculations.

(Of course, there are other questions that cannot be answered in such an effective treatment and that need to be adressed in a nonperturbative quantum theory of the underlying degrees of freedom)

 

[qsa]

http://arxiv.org/PS_cache/arxiv/pdf/1003/1003.2312v3.pdf

Statistical Origin of Gravity

Rabin Banerjee, Bibhas Ranjan Majhi†
S. N. Bose National Centre for Basic Sciences,
JD Block, Sector III, Salt Lake, Kolkata-700098, India

Abstract

Starting from the definition of entropy used in statistical mechanics we show that it is
proportional to the gravity action. For a stationary black hole this entropy is expressed as S = E/2T , where T is the Hawking temperature and E is shown to be the Komar energy. This relation is also compatible with the generalised Smarr formula for mass.