Thursday, October 13, 2005

Loops 2005

Yesterday, Cumrun gave a talk about the swampland in the Radcliffe Institute (which used to be a college for the girls only when Harvard was a male school, but eventually they merged). I liked it although it was not too new for many of us, of course. One of the points that Cumrun chose to emphasize was the derivation of the existence of dualities from the finiteness of the moduli spaces. Without the dualities, the moduli spaces (with the same metric, determined from the action) would have an infinite volume.

This approach shows how elegant and efficient some proofs of mathematical theorems may become in the future. You conjecture that an infinite sum or integral is miraculously equal to another sum or integral. You identify these objects as two different forms of a physical observable in string theory, and prove the required duality by finiteness of a moduli space which itself follows from the holographic principle. ;-) This kind of smart connection with complex calculus or even physics agrees with some of the sophisticated modern methods that were used in maths to prove various things.

These approaches are not incontroversial and several people asked their questions and expressed their opinion that it cannot lead to new insights about high-energy physics - maybe even "by construction". :-) I understand these comments. Everyone realizes that it is not guaranteed that new approaches to uncover the new wave of discoveries in theoretical physics will succeed. And most people realize that it is not easy today to find some really new and exciting stuff.

But whenever Cumrun or someone else says a sentence that is even partially wrong, someone immediately tries to correct him. We simply can't build and don't build on wrong assumptions and wrong statements. And if an assumption is uncertain, everyone acknowledges that it is uncertain. And if the assertion is ambiguous or incomprehensible, someone asks what it means. In summary, Cumrun and several colleagues try to identify the common patterns of the known vacua of string theory in order to learn some general lessons about quantum gravity; string theory is used as an experiment to study quantum gravity. These general lessons include
  • finiteness of the volume of moduli spaces in string theory
  • a special example of the previous point is the finiteness of the volume of the parameter space of CFTs with respect to the Zamolodchikov metric
  • upper bounds on the numbers of species, possibly correlated with the cutoffs
  • lower bounds on the numbers of species in various backgrounds
  • bounds on the strength of various interactions (under construction)
I totally understand Peter Woit who can't understand why this is an attempt to build on well-established and rigorous features of string theory and to find more general and far-reaching conclusions - with negative knowledge about string theory one simply can't do better. But there are fortunately other people in the world who must try to do better.

When I looked at Robert Helling's description of Loops 2005

plus the related papers and so forth, my impression was completely different than the impression from Cumrun's visionary talk. Their conference seems as a continuous inflow of ideas that are wrong and ideas that are not even wrong. Also, it seems that I am kind of familiar with the work of nearly all the speakers, and know why each of them is kind of incorrect (except Robbert Dijkgraaf who is usually right but who did not make it to the conference at the end).

Carlo Rovelli tried to derive the graviton propagator. We discussed this paper already. First of all, a graviton propagator only makes sense once we expand the metric around a background, which is in contradiction with the very general goals of loop quantum gravity. Second of all, in loop quantum gravity, it is known from the articles of Baez and others that singular simplices dominate the path integral, while Rovelli (incorrectly) assumes that the nice simplices that look like flat space give the most important contributions.

John Baez argued that there was some interesting progress in spin foam models, contradicting his previous but recent observations that there has been no progress in quantum gravity.

Lee Smolin proposed a model how macroscopic causality emerges (the problem in LQG is, first of all, that there is ultralocality and signals can't propagate at all for simple choices of the Hamiltonian). After Lee constructed quarks from LEGO, he conjectured that this model was connected with the Pioneer anomaly and low-angular-momentum discrepancies at WMAP. Well, maybe... ;-)

John Barrett presented a well-defined new version of a spin-network-like model which admits a diagrammatic expansion. I am pretty sure that these things can't be related to gravity, but the talk, if we believe Robert Helling, made sense.

Stefan Theissen tried to explain basics of string theory as known in the 1980s to our LQG colleagues.

Fotini Markopoulou argued that quantum gravity is like a self-correcting algorithm of a quantum computer. I can't say anything intelligent about such conjectured connections because I couldn't distinguish them from tens of crackpots' ideas flowing to my mailbox every week. It is plausible that there is something interesting going on in this direction - and it is appealing - but it just makes no sense to me so far.

Olaf Dreyer, possibly inspired by Lee, proposes to solve the problems that have been solved by decoherence and conjectures that quantum mechanics is not linear and probabilistic but non-linear and deterministic.

Other talks discussed modest questions about thermodynamics of discrete systems analogous to the Ising model; speculative connections about temperature and time evolution. There are several other ideas around but eventually Robert Helling stopped his reports because there was nothing in the talks.

Thanks, Robert, anyway! And a message for Peter Woit: if you think that Cumrun's program is comparable to the weird ideas from Loops 2005 listed above, then I must say that you are [the rest of the message has been censored to preserve our usual highest standards of politeness].

1 comment:

  1. Dear Lumo,

    Thank you very much for these comments, it is useful as always to hear your insights, whether they are right or otherwise.

    I don't know what is going on at the 'Not Even Wrong' blog. There was a comment on the latest thread by someone called Bryan, asking Lee Smolin whether the work on QFT being discussed got into causal non-perturbative mechanisms.

    The post has now been removed by Peter. It said something like this:

    1. The virtual charge of the vacuum surrounding the 'electron core' is polarised, attenuating the core charge by a factor of 137 when seen from a big distance.

    2. The Dirac solution for the magnetic moment of the electron, defined as 1 Bohr magneton, refers to this core charge.

    3. The Schwinger-Feynman first corrective coupling adds on a factor of 1/(twice pi times 137) giving 1.00116 Bohr magnetons, accurate to 6 significant figures (you need to take account of more couplings for greater accuracy).

    4. This addition of 1/(2.Pi.137) arises as follows. The virtual charges appear at a distance and at any one time, one virtual charge, say a virtual positron (which will be polarised slightly nearer the core than virtual electrons, due to electric attraction), associates with the real electron core. This association is presumably the Pauli exclusion principle. The virtual electron and real electron core are both spinning and orbiting (before the virtual electron annihilates with a virtual positron).

    5. Because of the orbit and two spins (this is the vague bit), the magnetism of the virtual charge adds only 1/(2.Pi) to the magnetism of the core (dipole moment of electron). In addition, there is a shielding factor of 137.

    6. The 137 shielding factor occurs because although magnetism is not attenuated by the polarised vacuum, the electric field is, and it is the electric field which is responsible for the polarisation of the vacuum, in turn resulting in the first coupling correction.

    This is quite a long explanation, but it is probably not entirely incorrect. The advantage of following this up may be the removal of renormalisation and generally obscure QFT mathematics, tidying up and ending in a unified theory of electricity and strong nuclear force.

    Because the electrostatic force of the fundamental particle cores is reduced by a 137 factor by the veil of polarised charge surrounding them, the core electrostatic force strength is equal in magnitude to the strong nuclear force!

    I can cite various research reports documenting this effect in high energy physics, for example experiments by Koltick and others in 1997 (P.R.L. paper) where the smashed electrons into positrons and observed a smaller coupling constant (I think 127 instead of 137) due to the partial penetration of the shield of polarised virtual charge around the core of the electron.

    Do you think that this quantum foam stuff should be endlessly ignored as bordering heresy on ether crackpotism, or do you agree that it is interesting to try to find a model for spacetime fabric which may allow us to calculate QFT without renormalisation?

    Best wishes,