Loops 07: Conference report (part 3, including discussion session)

Saturday 30/07

There were only two plenary talks this morning, followed by a discussion session. The first was by John Stachel, who is a specialist on philosophy and history of physics (with special reference to Einstein and relativity) . He introduced a general philosophy called "measurability analysis", which is based on analyzing and defining possible measuring processes and abstracting from them the quantities that need to be quantized (transformed into non-commuting operators). His analysis of GR suggests, to him at least, that the projective and the conformal structures of spacetime geometry are "what needs to be quantized" in quantum gravity. The second one was by Michael Reisemberger, who sketched with admirable clarity a canonical formalism for GR in which initial data are on two null intersecting hypersurfaces. The plus is that null initial data are free, not subjected to constraints. He provided a definition of the Poisson Bracket in this formalism and suggested that quantization leads to area discretization, though this is not yet solid ground.

In the discussion session, Carlo Rovelli followed the same procedure used in the Zakopane lectures and read some selected questions from a notebook that had circulated among the audience the previous days for people to write them. Obviously there were dozens of questions written and Rovelli, given the time constraints, had to select only a handful of them for people to discuss. These are the ones that made the cut:

1) Do we expect topology change in Quantum Gravity?

Oriti said that he expects a general framework to allow for topology change, but probably the classical limit can only be recovered on a sector that disallows it. Ashtekar was of the opinion that that canonical LQG framework must allow macroscopic topology change (if lots of spins become trivial, we have macroscopically a "branching" spacetime)

2) What is the relation of Quantum Gravity to the foundational questions in Quantum Mechanics?

Obviously, a question that provokes a lot of discussion. Thiemann and Rovelli are conservatives who think that QG and foundations of QM can be treated separately -for Rovelli's reasons see what he said in Zakopane. Bianca Dittrich thinks that we need to develop our understanding of relational observers. Lucien Hardy said that GR is at least as radical as QM, so it is unlikely that it can be treated with the standard QM framework. John Donoghue disagreed: according to him, effective field theory shows that GR breaks down at high energies, so the sensible thing is to modify it and keep QM. (As I said in a previous post, this overlooks the fact that in this context the what is meant by GR is not the exact Einstein theory, but the conceptual fact that spacetime is dynamical and not fixed.)

3) Could there be experimental consequences of fluctuating causal structure?

Sabine Hossenfelder mentioned the possible consequences for arrival time of photons, but stressed that this comes only from a phenomenological model with no relation to underlying theory. Hardy said that a "fluctuation", as a superposition between two classical states, would need some kind of interference experiment to observe, which is very difficult to be realizable in practice. I think Ashtekar got into a discussion with him here, but I couldn't follow it well enough to take notes -anybody remembers? Martin Reuter said that causal structure may be different for different observables used to probe it, and especially the scale of these observables.

4) What is finite in spin foam models?

Alejandro Perez gave a rather technical answer, of which the only notes I managed to take say: "some models (in 4D) are finite, some are not". Whoa, that's informative. Sorry.

5) Do we expect the fundamental theory to be combinatorial, or to be embedded in a pre-existing manifold?

Rovelli pointed out the conflict between Thiemann's new "Algebraic Quantum Gravity" approach, which is purely combinatorial, and Smolin's program to recover matter from graph braiding, which requires graphs to be embedded. Thiemann said that matter can be included in the algebraic approach, just as a part in the complete Hamiltonian. (Obviously, it would be more appealing if we could derive matter instead of putting it by hand -but can we?) José Antonio Zapata said that the basic thing we need to do is to understand how to build up a quantum theory on a differential manifold (one not previously equipped with a metric structure, I gather).

And now the last question. It asked, to all plenary speakers, to say they "dream for Loops '17"; that is, on their most optimistic possible view, what is the title and abstract of the talk they imagine themselves presenting within ten years?

Many of the answers were predictable and variations of a basic template: abstracts saying "we present a complete theory of quantum gravity with testable (or, in the most ambitious cases, confirmed) predictions." Ashtekar said something like this, adding that his estimated probability for this scenario was 0%. (But he also gave the in my opinion rather optimistic figure of 50% for the probability of having some experimental evidence to start resolving ambiguities.) Reuter had one of the most concrete dreams: "It is shown that LQG is equivalent to Asymptotic Safety, and that that the quantuization ambiguities in it are finite in number and equivalent to the dimensionality of the Non-Gaussian Fixed Point." And finally, there was an extremely amusing exchange between Thiemann and Alejandro Perez, which is a fitting conclusion to this series of posts:

Thiemann (reading his dream abstract): "We present quantum gravity corrections to the electron fine structure, and find that they are in agreement with experiments carried out by the author"

[laughter from the audience]

Perez (reading his dream abstract): "We show that Thiemann's calculations are totally wrong."

[hysterical laughter from the audience]