Friday, September 04, 2009

Is our Universe unique, and how can we find out?

If you have spare 45 minutes, here's a fun panel discussion from April 3rd, 2009, taken during the Origins Symposium at Arizona State University. If you click the O.S. link, you may find other panels with Brian Greene, Lawrence Krauss, Steve Pinker, and many others.

The discussion about the multiverse above was moderated by Paul Davies and there were five impressive personalities participating in it: Andrei Linde, Alan Guth, David Gross, Sheldon Glashow, and Alex Vilenkin. Two of them are Nobel prize winners, two of them may become ones in a few years, and the last one is not that infinitely far from them, either - sorry, Alex. ;-)
Update: Another similar ASU discussion about the origins of the Universe is here.
Your humble correspondent knows all of of them well - except for Andrei Linde whom I have only met a few times. You shouldn't resist to watch. ;-) Despite the significant differences in the opinions about the multiverse, it is a no-nonsense discussion. No irrelevant hippie crackpots or computer administrator crackpots are included in it.


Linde begins with some jokes and recalls the basic facts about inflation, including the ads arguing that it is the lunch where all dishes are served for free. While he is considering the anthropic arguments to be a self-evident truth, and he is mixing it with some cosmological facts and insights of string theory that have been really established, Gross who sits next to him is rightfully trying to decide whether he should smash Linde or have a heart attack instead. ;-)

Linde even argues that if we give up the multiverse, we would even have to do all the following three things: find alternatives to string theory, inflation, and a new solution to the cosmological constant problem. That's just ludicrous. Neither string theory nor inflation actually imply that there can't be a "single" Universe, and the anthropic "solution" to the cosmological constant problem has so little nontrivial evidence supporting it that the third condition is essentially equivalent to "find a solution because none is yet known, anyway".

Guth asks the question whether it is possible to define probabilities on an infinite multiverse, a question that has been repeatedly discussed - and mostly answered negatively - on this blog. He offers his own introduction to the cosmic inflation and the existing evidence supporting it. Alan debunks some negative noise about "anything goes" created by his subpar, arrogant colleague Lawrence Krauss.

Guth thinks that "almost all" models of inflation lead to eternal inflation - which consequently produces a multiverse. Well, that's nice but the question whether it makes any physical sense to "continue time" through the nucleation of a bubble is very controversial. Nothing can be "remembered" about the parent universe in this process, so it is questionable whether the "pre-history" exists in a physical sense.

Do we need probabilities in such a multiverse? Yes, he says, because everything that can happen will happen somewhere. Well, I agree - I just completely disagree with the way how the anthropic people calculate these probabilities. At least, Guth realizes that (uniform) probabilistic distributions on infinite sets are ill-defined: he gives an example.

But this stuff is completely irrelevant because probabilities of any local enough phenomenon are only determined by applying the rules of quantum mechanics to the initial state - the counting of "copies" in the multiverse is inconsequential because all of these copies in multiverses with similar enough effective field theories will follow the same probabilistic predictions! Sadly, Alan Guth thinks that some models predict "Boltzmann Brains". Only a wrong way of calculating the odds can make such predictions.

Vilenkin paints the multiverse as colorful balls embedded within each other. He correctly says that whether or not you like it, we may be in a multiverse which would reduce our ability to predict. Even though the other universes are behind the cosmic horizon, the theory is testable, he says, and he refers to probability distributions (obtained by the "relative fractions" of observers or galaxies within the multiverse, the same wrong anthropic formula).

Vilenkin defends the principle of mediocrity - which is also a principle of a mediocre approach to cosmology. ;-) The cosmological constant is presented as the only working prediction of these anthropic predictions.

Gross says that the title of his talk was pre-determined by Davies: Is the Universe just right for life? Many answers have been given during the last millennium but now the question has become a part of science. Gross is going to be negative about the anthropic reasoning. First, he sketches it, with a nice picture of the stringy landscape where most places suck. One needs a landscape - arising in string theory - but also a mechanism to populate them - offered by inflation.

But... in string theory, these vacua are really metastable, uncontrollable, not even exact solutions that are not usable as full-fledged cosmologies. The mechanism to populate them - eternal inflation - is technically and even conceptually shaky, because of the causal disconnect between them, the ill-defined measures, and other problems.

Moreover, it's not clear whether such things follow from string theory because we don't know what it inherently is: string theory should exchange the names with the Standard Model (Weinberg's fault!). ;-) A "framework" would be better. In this sense, it's on par with the effective gauge theory - a different parameterization of a similar class of models. It will inevitably be with us forever because it is continuously connected with the Standard Model. (I am not making comments simply because I agree with every word he says.)

Something fundamental is missing. It's the first time we are constructing a theory of the spacetime that should include the "whole thing". That's why we need to construct a full cosmology, and we need the full cosmological standards for something to be called a solution. The right "solution" may even require us to understand the laws for the initial conditions which we don't yet know.

The dangers of the anthropic principle is that it is hard or impossible to falsify and often tautological. And it goes against the history of physics which has been reducing the number of necessary parameters for quite some time. It's an easy way out. And it thrives on ignorance: the more you don't know something, the more you believe the anthropic principle.

Gross once asked Davies: how can we disprove the anthropic principle? Isn't it necessary to calculate everything? And Davies said: Yes. Well, it's hard but Gross replied: We will try! ;-) Because we don't know the final theory of quantum cosmology and of the vacuum selection, we shouldn't give up this goal just because of one number, the cosmological constant. Just because we don't yet know the solution doesn't mean that the solution doesn't exist - a logical trap we must avoid.

After all, it's just 64 orders of magnitude below the generic expectations in a sensible SUSY theory, and the typical mass scale - the fourth root - is just 16 orders of magnitude too small which is not unprecedented. Similar numbers have been explained.

For example, Dirac wanted to explain why the proton mass was 19 orders of magnitude below the Planck mass. Dirac didn't invoke the anthropic principle: good for him. He suggested that it was linked to other big and small numbers in physics. Dirac made predictions and his theory could have been falsified. But the right explanation was found in QCD: the strong coupling is 1/25 or so at the Planck scale - reasonable - and by the log running, it has to become comparable to one at a much lower (QCD) scale. Similarly, the GUT/electroweak hierarchy is explained logarithmically assuming SUSY while light neutrinos arise via the seesaw mechanism.

No one would pay attention to the anthropic principle if we could calculate the cosmological constant. The main problem is not that it is too small: the main problem is that we don't yet know the mechanism. So we will figure it out. The strategy is to observe, experiment, and calculate.

Glashow presents himself as a black sperm and proves that this is the best universe among all of them because we can observe it. ;-) New connections will be found in real time, real space, at home. Glashow complained about the SSC cancellation and the drought that has been around for 17 years. But things are getting exciting - the LHC, astrophysical observations etc. Glashow says several confusing things about the observation of dark matter and sketches plans for the LIGO, ILC, and other experiments. Soon, we will see things again so the excitement will become more down-to-Earth, he thinks. Glashow is charming but his brief monologue was somewhat shallow and off-topic concerning the merits.

No comments:

Post a Comment