Carlip's talk at Berkeley string seminar (featured results CDT et al)

[marcus]

The main weekly seminar at the UC Berkeley physics department's Center for Theoretical Physics is the string seminar---or one of a couple anyway---I think Petr Hořava is the main person in charge. His office is right by the seminar room. Whole setup is pretty nice.
Building's been remodeled.

Today they had a visitor from UC Davis come and give a talk. Steve Carlip. He drew on results from Loll Triangulation QG, Loop QG, Reuter AsymSafe QG, and other 4D approaches---results about spontaneous decline of dimensionality from 4D down to around 2D at very tiny scale.

There were about 16 of us in the audience---most looked like postdocs or advanced grad students. Hořava, who is still fairly young too, appeared to be the most authoritative person in the audience and asked quite a lot of questions.

Hořava is pronouced "Ho zha va" because of the mark over the letter ř.

Carlip's talk was similar to what he gave at the Planck Scale conference around start of July, at Wroclaw. It was the opening talk of the conference. The video is online. It's a good talk about a very interesting possibility of what could happen at small scale in nature.
http://www.ift.uni.wroc.pl/~rdurka/planckscale/index-video.php?plik=http://panoramix.ift.uni.wroc.pl/~planckscale/video/Day1/1-1.flv&tytul=1.1%20Carlip
Might not of course, but could.

Pointed to by these different 4D QG approaches all coming to unexpectedly similar conclusions from different directions by different reasoning. You imagine a 4D block universe generated by one of these approaches, and think of a diffusion process or random walk starting at some point and see how the probability of accidentally returning to start, after going a path of length s, depends on s. The probability trails off kind of as 1/s^d/2.

The bigger the dimensionality d, the faster you get lost, and the less likely you are to accidentally find your way back home. So if you are simulating quantum universes in a computer you can "empirically" measure the dimensionality by having the computer run a random walk in the quantum spacetime you generated.
Well that is how Loll's Utrecht group did it.

Interestingly enough Carlip seems to like the CDT approach and he has a Phd student named Rajesh Kommu duplicating Loll's results but with their own UC Davis code.
It is significant that they didn't just take the Utrecht CDT code. They built their own CDT Monte Carlo from the ground up. That is a way of being sure that the computer "experiment" of generating these small quantum universes is repeatable.

CDT is also being run by Joe Henson at Perimeter, using a large computer there, but I don't know how that is going. Maybe somebody knows?

 

[marcus]

Anyway some points to make:

It was a good sign that the UC Bekeley socalled "string" seminar would host a mainly 4D QG talk. And notice that Horava (star faculty, former young co-author with Witten,... etc.) is now mainly known for his own 4D QG method, called Horava-Lifgarbagez gravity. This in itself shows flexibility and border-crossing ability. I think it is part of this trend for the lines to fuzz, divisions to fade, for people to cross over, it's all to the good and I wish more of that kind of thing was happening!

As I recall both Loll and Reuter were invited plenary speakers at the Loops 2005 conference four years ago. Loop researchers were very interested to hear about CDT and AsymSafe then. And both approaches already had published their 4->2 spontaneous dimensional reduction results at that time. That has had increasing impact in the subsequent 4 years.

Carlip basically had two main things to say. One is to argue that in the absence of experimental guidance where there is no clear right QG and all theories so far may be wrong, it's reasonable to pay some attention to a feature that shows up in several different approaches. He gave the example of black hole entropy--e.g. the relation to area. Several different approaches arrived at common or overlapping conclusions about black hole thermodynamics. So we tend to take that more seriously than if only just one approach came up with the idea that area and entropy are connected.

The other thing was more theoretical and took up the second half of the talk. That is how you can use some classical GR solutions to get intuition about how this dimensional reduction comes about, and visualize it. Even though you don't have the quantum mechanical blessing on it, you can still picture a classical Kasner or "mixmaster" mess occurring as you plunge down and down in scale---as you zoom into Planck scale and beyond. I am not giving a fair description but you can check out the Wroclaw Planck Scale conference video of Carlip's talk.