Thursday, December 07, 2006

Joe Polchinski, one of the key physicists behind the second superstring revolution and the main father of D-branes, agreed to review two recent notorious physics books for American Scientist. And I would say that he did so in the characteristic old-fashioned perfectionist Polchinski fashion. ;-)

His 25 kilobytes of text or so describe some of the most universal errors, misconceptions, and misleading statements in these two books, together with Joe's own viewpoint on many of these broader questions about physics and together with some kind words. It is an interesting reading.

You can also read a technically expanded version of Joe's article at

Joe Polchinski became a guest blogger with one of our potential competitors. No doubt, he has added some fuel to the controversy as some people call it. But when the fuel is added properly, it can sometimes burn the garbage into water and pure carbon dioxide which is a friendly gas that we call life. Let's hope that this environmentally friendly scenario will work out in this case. ;-)

Below, you find links to the paperback edition of Joe's book.

.....

Polchinski praises Smolin's description of the enthusiasm during the first superstring revolution. Then he shows why Smolin's statement that string theory predicted that the cosmological constant couldn't be positive is the first big myth of that book: the Maldacena-Nunez paper requires a special discussion at this point. Joe remembers the assumptions that went into their theorem and the fact that everyone always realized its limitations.

Polchinski continues by debunking various strange and clearly incorrect constructs of Smolin about holography such as "unpredictive" Maldacena's correspondence or "non-quantum" Maldacena's correspondence or "weak" Maldacena's correspondence and many others: much like I have explained in the review of Smolin's book (remember my analogy with the weak Darwinism?), Polchinski also clarifies that it is not possible to define any "weaker" version of Maldacena's duality that wouldn't be falsified by the known tests. More concretely, the conjecture crucially depends on quantum mechanics and it cannot break beyond the classical limit in any sense.

Polchinski stresses that physics works differently than mathematics and must avoid the temptation of rigor. He also emphasizes another well-known fact for the Reference Frame readers, namely that physics of string theory satisfies background independence even if its particular formulations don't make it manifest. String theory is much more background-independent than the existing forms of loop quantum gravity, he writes. In AdS/CFT, the boundary of the spacetime is fixed but the interior can fluctuate in an arbitrary way (including the creation and evaporation of black holes) - the whole bulk and its diffeomorphism symmetry is emergent - which is another way to falsify Smolin's weird and frequently used speculation that string theory isn't background-independent.

Polchinski mentions that there is nothing wrong with the evolution of the mathematical formalism that describes the same physics and quotes Einstein's clumsy formalism behind special relativity, replaced by Minkowski's modern notation, as another example besides string theory.

Joe describes that our situation with the non-perturbative definition of string theory is in some sense better than the situation of quantum field theory that has lasted for 50 years and it is better than Woit's opinions: we now have a full non-perturbative definition for many superselection sectors of string theory which became accessible because of the duality revolution, a fact that is apparently unfamiliar for both authors. Joe gives some examples of such definitions in a footnote - superrenormalizable theories defining some lower-dimensional AdS backgrounds.

Polchinski explains that the original Planck scale (something that Smolin calls a lie) is an upper bound on the scale where quantum gravity has to become relevant but the actual fundamental scale may be lower. On the other hand, it can't be arbitrarily low because of many constraints. This explains why it is difficult to test quantum gravity and why this fact can't be used as a criticism of a particular theory. Joe discusses the multiverse as a potentially important prediction of string theory if it is proven; this blog hopes it will not be ;-) although we of course appreciate the strength of the recent technical results that support Joe's attitude. See also Joe's review of the cosmological constant problem.

Some paragraphs are dedicated to RHIC and AdS/QCD. Joe explains, much like we did, that you can't dismiss this field as a decoupled mathematical spinoff of string theory because Nature recycles many of the mathematical structures at various places and the quantum gravity dual to the RHIC phenomena is therefore fundamentally the same theory of gravity as the theory of gravity at the fundamental scale. For example, in the vacua with warped geometry, the QCD gravity is just another part of the geometry shared by the fundamental scale gravity.

The author of the less interesting and, using Joe's words, more "repetitious" book among the two books agreed that Polchinski's review debunking the two books is thoughtful. But he didn't resist and repeated many of his usual crackpot concepts together with some rumors about Larry McLerran's opinions about holography. Rumors about jokes presented by an old physicist at a conference may be interesting but as Jeff Harvey has pointed out, if you look at the actual paper(s) by McLerran et al., they call the description of the RHIC viscosity via AdS/CFT an "amazing theoretical discovery" in the last paragraph of page 2 of their paper.

Joe continues by debunking the conspiratory sociological theories in both books. While he says that sociological effects do exist, he explains that our field contains contrarians and mavericks ;-) and that it is not disconnected from the rest of science because people are crossing the fuzzy boundaries all the time: the insiders and outsiders in various subfields and adjacent fields constantly interact. He re-iterates the idea about the market of ideas, previously mentioned by Aaron Pierce, Brian Greene, and Barton Zwiebach: the theorists would have already migrated if they didn't think that string theory is by far the most promising approach to these questions while the actual brain drift seems to move in the opposite direction. Also, he says that many statements that the authors claim to be sociological in nature are actually artifacts of their different (i.e. flawed) technical reasoning.

In the final footnote, Joe suggests that various Smolin's favorite theories of Lorentz violation are unlikely to be compatible with the accurate experimental tests of Lorentz symmetry because the loop effects would amplify the Lorentz violation at long distances. Similar issues were discussed in objections to loop quantum gravity. Polchinski suggests that these calculations should have already been made.

Throughout the text, Polchinski argues that we have good reasons to think or know that we are discovering something important instead of inventing it. And the unexpected convergence of all these originally different ideas seems to be a highly non-trivial argument in favor of string theory and its validity.

Finally, Polchinski recommends better books by Greene, Susskind, Randall, and Vilenkin because the books of Woit and Smolin fail to capture the logic and spirit of string theory.

And that's the memo: I almost forgot that it is technically Polchinski's memo, not mine. ;-)

Nullity

Clifford Johnson has pointed out that the intelligent journalists at BBC have announced that after 1200 years of efforts, the problem how much is 0/0 has been finally solved. Dr. James Anderson has figured out that the result is nullity: the symbol is a skiing Mexican penguin. I agree with Clifford that this looks exactly like an article from The Onion. ;-)