Thursday, March 06, 2014 ... Français/Deutsch/Español/Česky/Japanese/Related posts from blogosphere

Brian Greene's talk on the state of string theory

Stephen and Vincent Della Pietra – who are not Capo di tutti capi because there are two of them; instead, they are fratelli – donated a few million dollars to Stony Brook and launched their lecture series. Recent speakers included (or the coming one will include) Wilczek, Linde, Veltman (and Schwarz).



In October 2011, Brian Greene gave the talk on "The State of String theory" which was finally posted to YouTube and if you can sacrifice 76 minutes (or a part of them), you are invited to watch the talk.




There are lots of the usual and some unusual introductory comments related to string theory – its history, major conflicts in physics, small vs long scales, why the theory unifies, pants diagrams, extra dimensions and their physics.

Since 27:00, it's more about the "present" – how to extract phenomenology, nonperturbative formulations including the Matrix Theory Hamiltonian, AdS/CFT, dualities, M-theory, landscape, experimental tests, braneworlds, cosmology, inflation, singularities, impacts on enumerative geometry, quantum geometry, a report card, emergence and holography.




Questions begin at 1:03:40 and Brian's answers are often amusing. They suggest a significant gap between the diplomatic formulations he often likes to offer to large audiences and what he really thinks on the other hand. With this variability, the hypothesis that he actually thinks the same thing about most of these issues (and the non-orthodox "interpretations" of quantum mechanics unfortunately don't belong to this list) as your humble correspondent is totally viable. I also have some spread of the "degree of diplomacy" depending on the context, it's just visibly smaller than Brian's.

The first man complains that Brian omitted the competitors. Brian says that he didn't see any competitors. People laugh and he says that it is a fair question and switches to the mode of Brian Greene the diplomat. He enumerates the guys who are doing loop quantum gravity and how much they believe that they have an alternative and so on. He wraps this flattering discussion by saying that it's good that people are working on various alternatives – and he's also happy that he personally doesn't work on that LQG pile of crap! ;-)

Brian offers an even more entertaining two-colored answer to a question about the "superluminal OPERA neutrinos" that were hot at that time. He says how smart these people are and how they have surely incorporated all the effects of the GPS synchronization, slowdown of light in the air, turbulence of the air, [he enumerates about 10 other possible sources of errors]. These people are so careful, you know. Brian keeps the consistently diplomatic language and says that he might still need some truly independent measurement to be compelled. The punch line arrives in his last sentence: "In fact, I would bet anything I hold dear that their result is wrong." The audience explodes in laughter because in the context, the sentence is a work of a comedian. Recall that the bogus "superluminal" OPERA result was an artifact of a loosely connected optical cable.

The third question is about the links of string theory and the Higgs boson. Brian says some of the same things I did but he also discusses the (currently indeed emerging) nightmare scenario (Higgs and nothing else at the LHC) and the reactions of the funding agencies to this "surprising and exciting" possibility.

When Greene is asked the overly popular question whether the landscape makes string theory unfalsifiable, Brian says that there are different ways to deal with this question. The first way, he shows, is offered by an "offensive cartoon". Just to be sure, the answer to the question is "PAK, you keep on talking like a bitch so I'm gonna slap you like a bitch". I have seen and given many answers to the question but I still think that this is the most accurate and appropriate one. Brian says that "I don't even want to show that [answer]" but I suspect he also thinks it's the best answer on the market. He says it's "not his perspective at all, it is completely inappropriate", so he offers another, less compelling answer. He points out it is not obvious that you can get anything out of the many vacua. The points are discrete and sparse, so after some measurements, you may produce predictions. Second, he says that the statistical properties of the vacua should be studied – clusters, groups, statistical predictions etc. He says that the problem is that it is hard to make it. I think that a more conceptual problem is that one can only make predictions if he has some probability measure and there's no known natural probabilistic distribution on the space of vacua (the egalitarian one is clearly wrong).

What is the difference between the choices of fields and parameters of the Standard Model on one side and properties of the compactifications in string theory on the other? Well, the former ones are continuous numbers etc.; the latter are just some discrete data. So the latter yield no undetermined, adjustable continuous dimensionless parameters. For practical applications – whether we really can say what's right or wrong – the framework of string theory is as predictive or unpredictive as the framework of quantum field theory.

Michael Douglas adds the last comment – a more careful analysis of early cosmology in string theory may actually tell us what the shape of extra dimensions ultimately looks like or prefers to look like which may make predictions possible. Brian Greene says "exactly" and that's the end of the talk.

Add to del.icio.us Digg this Add to reddit

snail feedback (0) :