Classical theories and AdS/CFT

Question

When I was editing the Physics.SE tag wiki for ads-cft, I initially wrote something on the lines of :

The AdS/CFT correspondence is a special case of the holographic principle. It states that a gravitating theory in Anti-de-Sitter (AdS) space is exactly equivalent to the gauge theory/Conformal Field Theory (CFT) on its boundary.

Then I thought, "Wait, it can't be ANY gravitating theory, right? It must be a theory of Quantum Gravity, right?". So I changed it to:

The AdS/CFT correspondence is a special case of the holographic principle. It states that a quantum gravitating theory in Anti-de-Sitter (AdS) space is exactly equivalent to the gauge theory/Conformal Field Theory (CFT) on its boundary.

Now, I'm a bit unsure. Can AdS/CFT work for classical gravity theories? .

I have seen the questions Which CFTs have AdS/CFT duals? and What is on the AdS side in AdS/CFT supergravity or string theory?, but my question is a bit different.. It is a special case of the opposite of the first question, and is more general than the second. From Lubos Motl's answer to the second question, I see that supergravity theories don't form AdS theories with CFT duals, but is that true for all classical theories?.

And a bit more general question: If the answer to the above question^ is yes, then do all gravitational theories in AdS's have CFT duals? . This is pretty much the opposite of Which CFTs have AdS/CFT duals?.

For example, is General Relativity such a theory; with a CFT dual? If so, what would be it's CFT dual? What about... Nope I'm not going to ask about Newtonian Gravity, or Aristotilean Gravity, for obvious reasons. And certainly not LQG (something that doesn't respect holography in the most trivial situations, couldn't here.).

This post imported from StackExchange Physics at 2014-03-07 16:31 (UCT), posted by SE-user Dimensio1n0

Comments

Every credible example of AdS/CFT that I can think of, ultimately identifies a quantum field theory (the CFT) with a string theory (or M theory) in an AdS space (ie AdS_k x M, where M is some manifold). If people say that the duality relates a gauge theory to quantum supergravity, classical supergravity, classical gravity... they are just talking about different limits of the string theory, which is the true ultimate object on the other side of the duality.

This post imported from StackExchange Physics at 2014-03-07 16:31 (UCT), posted by SE-user Mitchell Porter

---

There are various generalizations of AdS/CFT like "Kerr/CFT" or a "hydrodynamic" version which don't have a clear string-theory parent, but you would expect that in their final form, these dualities also would have quantum gravity and thus(?) string theory on the gravity side.

This post imported from StackExchange Physics at 2014-03-07 16:31 (UCT), posted by SE-user Mitchell Porter

---

@MitchellPorter: Thanks, but are these "generalisations" exact too, like AdS/CFT? Or do they become exact only when you're out of the classical limit, i.e. when they are in their final form and have quantum gravity.

This post imported from StackExchange Physics at 2014-03-07 16:31 (UCT), posted by SE-user Dimensio1n0

---

The attitude I would adopt, is that every such correspondence has a quantum/quantum exact correspondence as its ultimate foundation. But I can't prove it; I'm just assuming that the situation in AdS/CFT is reproduced in these other areas, when we finally know all the facts...

This post imported from StackExchange Physics at 2014-03-07 16:31 (UCT), posted by SE-user Mitchell Porter

---

The "fluid/gravity correspondence" (which is the hydrodynamic one) arxiv.org/abs/1107.5780 is a classical/classical correspondence. The Kerr/CFT correspondence arxiv.org/abs/1203.3561 I think just relates some macroscopic properties of the black hole (like charge, mass, angular momentum) to basic parameters of the CFT (e.g. "central charge"). This is ongoing research so we don't have the full context for these relationships yet.

This post imported from StackExchange Physics at 2014-03-07 16:31 (UCT), posted by SE-user Mitchell Porter

---

@Dimension10: The correct question would be : Does holography works for classical gravity ?. The fact that the entropy of a black hole is proportionnal to its surface suggests that yes. But it does not mean that there is a quantum field theory (defined on the horizon) dual to a non-quantum gravity theory. If you take a limit like (weak energy, non supersymmetric, non quantum), I think the limit has to be taken in both sides on the holography.

This post imported from StackExchange Physics at 2014-03-07 16:31 (UCT), posted by SE-user Trimok

Answer 1

In principle, the AdS/CFT correspondence relates a conformal quantum field theory to a quantum theory of gravity (string theory). The key to the solution of all this confusion can be found in taking appropriate limits. It turns out that if you have strong coupling on the string theory side, you have a weakly coupled CFT and vice versa. The weakly coupled limit of string theory is classical (super-gravity), which now corresponds to a strongly coupled conformal quantum field theory. This is one of the main reasons why the correspondence is interesting: it enables one to use perturbative string techniques in order to solve strong coupling field theory problems. There are many efforts to apply this to QCD (which is a strongly coupled field theory), with remarkable success.

But how is this consistent with the assertion that fluid/gravity duality is a classical/classical duality?

Classical on the field theory side in this context does not mean that the underlying theory is not a quantum field theory, it definitely is a strongly coupled QFT. However, in the fluid/gravity duality, the long-wavelength limit is used, which allows one to formulate the problem in terms of the classical Navier-Stokes equation. One can now use the weakly coupled gravity-side to determine parameters for fluid-dynamics.

For a good introduction to the matter at hand, see both http://arxiv.org/pdf/0905.4352v3.pdf and http://arxiv.org/pdf/0712.0689v2.pdf .

This post imported from StackExchange Physics at 2014-03-07 16:31 (UCT), posted by SE-user Frederic Brünner

Comments

Thanks a lot Frederic Brunner for this answer!