Monday, September 28, 2015

No, Loop Quantum Gravity has not been shown to violate the Holographic Principle

Didn't fly.
Tl;dr: The claim in the paper is just wrong. Read on if you want to know why it matters.

Several people asked me for comments on a recent paper that appeared on the arxiv, “Violation of the Holographic Principle in the Loop Quantum Gravity” by Ozan Sargın and Mir Faizal. We have met Mir Faizal before; he is the one who explained that the LHC would make contact to parallel universes [spoiler alert: it won’t]. Now, I have recently decided to adapt a strict diet of intellectual veganism: I’ll refuse to read anything produced by making science suffer. So I wouldn’t normally have touched the paper, not even with a fork. But since you asked, I gave it a look.

The claim in the paper is that Loop Quantum Gravity (LQG), the most popular approach to quantum gravity after string theory, must be wrong because it violates the Holographic Principle. The Holographic Principle requires that the number of different states inside a volume is bounded by the surface of the volume. That sounds like a rather innocuous and academic constraint, but once you start thinking about it it’s totally mindboggling.

All our intuition tells us that the number of different states in a volume is bounded by the volume, not the surface. Try stuffing the Legos back into your kid’s toy box, and you will think it’s the volume that bounds what you can cram inside. But the Holographic Principle says that this is only approximately so. If you would try to pack more and more, smaller and smaller Legos into the box, you would eventually fail to get anything more inside. And if you would measure what bounds the success of your stuffing of the tiniest Legos, it would be the surface area of the box. In more detail, the amount of different states has to be less then a quarter of the surface area measured in Planck units. That’s a huge number and so far off our daily experience that we never notice this limit. What we notice in practice is only the bound by the volume.

The Holographic Principle is a consequence of black hole physics, which does not depend on the details of quantizing gravity, and it is therefore generally expected that the entropy bound must be obeyed by all approaches to quantum gravity.

Physicists have tried, of course, to see whether they can find a way to violate this bound. You can consider various types of systems, pack them as tightly as possible, and then calculate the number of degrees of freedom. In this, it is essential that you take into account quantum behavior, because it’s the uncertainty principle that ultimately prevents arbitrarily tight packing. In all known cases however, it was found that the system will collapse to a black hole before the bound is saturated. And black holes themselves saturate the bound. So whatever physicists tried, they only confirmed that the bound holds indeed. With every such thought-experiment, and with every failure of violating the entropy bound, they have grown more convinced that the holographic principle captures a deep truth about nature.

The only known exception that violates the holographic entropy bound are the super-entropic monster-states constructed by Hsu and collaborators. These states however are pathological in that not only will they inevitably go on to collapse to a black hole, they also must have come out of a white hole in the past. They are thus mathematically possible, but not physically realistic. (Aside: That the states come out of a white hole and vanish into a black hole also means you can’t create these super-entropic configurations by throwing in stuff from infinity, which should come as a relief to anybody who believes in the AdS/CFT correspondence.)

So if Loop Quantum Gravity would violate the Holographic Principle that would be a pretty big deal, making the theory inconsistent with all that’s known about black hole physics!

In the paper, the authors redo the calculation for the entropy of a particular quantum system. With the usual quantization, this system obeys the holographic principle. With the quantization technique from Loop Quantum Gravity, the authors get an additional term but the system still obeys the holographic entropy bound, since the additional term is subdominant to the first. They conclude “We have demonstrated that the holographic principle is violated due to the effects coming from LQG.” It’s a plain non-sequitur.

I suspect that the authors mistook the maximum entropy of the quantum system under consideration, previously calculated by ‘t Hooft, for the holographic bound. This is strange because in the introduction they have the correct definition for the holographic bound. Besides this, the claim that in LQG it should be more difficult to obey the holographic bound is highly implausible to begin with. LQG is a discretization approach. It reduces the number of states, it doesn’t increase them. Clearly, if you go down to the discretization scale, the number of states should drop to zero. This makes me think that not only did the authors misinterpret the result, they probably also got the sign of the additional term wrong.

(To prevent confusion, please note that in the paper they calculated corrections to the entropy of the matter, not corrections to the black hole entropy, which would go onto the other side of the equation.)

You might get away with the impression that we have here two unfortunate researchers who were confused about some terminology, and I’m being an ass for highlighting their mistakes. And you would be right, of course, they were confused, and I’m an ass. But let me add that after having read the paper I did contact the authors and explained that their statement that the LQG violates the Holographic Principle is wrong and does not follow from their calculation. After some back and forth, they agreed with me, but refused to change anything about their paper, claiming that it’s a matter of phrasing and in their opinion it’s all okay even though it might confuse some people. And so I am posting this explanation here because then it will show up as an arxiv trackback. Just to avoid that it confuses some people.

In summary: Loop Quantum Gravity is alive and well. If you feed me papers in the future, could you please take into account my dietary preferences?

20 comments:

Uncle Al said...

Surface area defining volume sounds like dimensionality. "I am an ass" Judges 15:15-17. Mucking stalls theory is not ponies,

http://www.abarim-publications.com/Meaning/Ramath-lehi.html

"I’ll refuse to read anything produced by making science suffer" Newton fumbled c, h, and k_B (GR, QM, statistical thermodynamics) with no derivation errors. Given baryogenesis (excess of matter over antimatter), identify and repair the postulate enforcing "no." Falsification offends physics as Bolyai offends Euclid. "physically realistic 90 day geometric Eötvös experiment; one day enantiomers' microwave rotational temperature spectra, Newtonian apple or road apple. Look.

matrix said...

Not related with this, but recently I read http://www.dailymail.co.uk/sciencetech/article-2769156/Black-holes-NOT-exist-Big-Bang-Theory-wrong-claims-scientist-maths-prove-it.html. I'm a physics enthusiast, but do not have the knowledge to dig dipper. Have you heard of it, or is it just some media BS?

Sabine Hossenfelder said...

matrix: Laura Mersini-Houghton's claim is wrong. She is simply using the wrong equations. I wrote about this here. I find it extremely depressing to see these things propagated in the media without even the least amount of scrutiny. Best,

B.

matrix said...

Thanx a lot for your reply, Best, A

kashyap vasavada said...

Even ignoring the authors' arguments, my difficulty with LQG is that it violates Lorentz invariance (well before Planck energy?).Is this right and have you written about this problem?

andrew said...

"But let me add that after having read the paper I did contact the authors and explained that their statement that the LQG violates the Holographic Principle is wrong and does not follow from their calculation. After some back and forth, they agreed with me, but refused to change anything about their paper, claiming that it’s a matter of phrasing and in their opinion it’s all okay even though it might confuse some people."

This would sound a lot more sincere if it didn't go to the very title of the paper.

Sabine Hossenfelder said...

andrew: yes... and it appears in various places in the paper as well, together with an explicit (and correct) definition of what they mean by holographic principle. At the very least they should have corrected the awkward grammar in the title ;)

Sabine Hossenfelder said...

kashyap,

Yes... Let me just vaguely say the problem is unsolved and for what I can tell the LQG people can't agree on just what happens to Lorentz-invariance. It was recently proposed that to solve the issue with Lorentz-invariance in LQG, one needs matter-type interactions that are similar to those in string-theory, see my earlier post. Best,

B.

A. Mikovic said...

I don´t see why a QG theory has to satisfy the Holographic Principle (the maximal entropy of a compact region of space should be bounded by its area, according to Wikipedia). The only thing one should require is that the entropy of a black hole should be dominated by the horizon area for large areas. AdS/CFT conjecture is cited as a realization of the Holographic Principle, but this is something specific to string theory. One can have a QG theory which has nothing to do with string theory, for example, spin-foam models or Casual Dynamical Triangulations.

Sabine Hossenfelder said...

A. Mikovic,

One can debate this point, and I have discussed this here for example. I didn't want to bring it up here because it would merely have distracted from the main point, which is that the argument in the paper is plainly wrong, regardless of whether you thought it was important even if it had been correct. Best,

B.

Waterbergs said...

Love it Sabine. I am thinking of adopting your diet myself. Should definitely avoid some intellectual indigestion issues.

P

stor said...

What exactly is the problem with Lorentz invariance?

Sabine Hossenfelder said...

It isn't really understood how Lorentz-invariance is recovered in the loop quantization, if it is modified and if in which way. The finding that one has a minimum area and volume seems to clearly violate Lorentz-invariance. But then these are the lowest quantum numbers for the spectrum of some operators, and it's not clear what happens to actual observables.

stor said...

That's what I don't understand, why minimum area "clearly" violates the invariance? In some quantum mechanical situations there is a minimum energy or angular momentum (or other observables with discrete and bounded from below spectrum), but there is no problem with translational or rotational invariance.

Sabine Hossenfelder said...

Stor,

Yes, in a nutshell the angular momentum analogy is Carlo Rovelli's argument. I am not sure exactly which quantum mechanical system you have in mind. Of course you can have quantities of dimension energy, length, area, etc that are invariant. That's different from saying it's an actual energy, length, area (of something). Thus the question of observables. Best,

B.

stor said...

Wait, I still don't get it. What is wrong with Rovelli's analogy/argument?

Sabine Hossenfelder said...

Stor,

I never said the argument is wrong. I said that without knowing what's an observable and how to calculate this and how it transforms under Lorentz-transformations one doesn't know whether it's wrong.

stor said...

Ah, ok, I thought you said that quantized volume, area, length clearly shows that there is a problem with Lorentz invariance.

Sabine Hossenfelder said...

No, sorry if it came across this way. What I mean is, it is a potential problem that needs a good answer, and not just a "it might be like this" kind of argument.

stor said...

No, no, I misunderstood what you'd written.