Not Even Wrong

There’s almost too much to keep track of the last couple days on the string theory controversy front:

Burton Richter of SLAC has a Reference Frame piece in the latest Physics Today entitled Theory in particle physics: Theological speculation versus practical knowledge. Richter shares my point of view that the Landscape studies currently popular in string theory are not science:

To me, some of what passes for the most advanced theory in particle physics these days is not really science. When I found myself on a panel recently with three distinguished theorists, I could not resist the opportunity to discuss what I see as major problems in the philosophy behind theory, which seems to have gone off into a kind of metaphysical wonderland. Simply put, much of what currently passes as the most advanced theory looks to be more theological speculation, the development of models with no testable consequences, than it is the development of practical knowledge, the development of models with testable and falsifiable consequences (Karl Popper’s definition of science)…

The anthropic principle is an observation, not an explanation… I have a very hard time accepting the fact that some of our distinguished theorists do not understand the difference between observation and explanation, but it seems to be so…

What we have is a large number of very good people trying to make something more than philosophy out of string theory. Some, perhaps most, of the attempts do not contribute even if they are formally correct.

The issue of Nature that just came out today has an article about the controversy by Geoff Brumfiel with the title Theorists snap over string pieces: Books spark war of words in physics. He describes Lubos Motl’s reviews of the Smolin book and mine on the Amazon web-site, and quotes Polchinski and Susskind. The reaction of string theorists to the books is said to be:

Few in the community are, at least publicly, as vitriolic as Motl. But many are angry and struggling to deal with the criticism. “Most of my friends are quietly upset,” says Leonard Susskind, a string theorist at Stanford University in California.

and

The books leave string theorists such as Susskind wondering how to approach such strong public criticism. “I don’t know if the right thing is to worry about the public image or keep quiet,” he says. He fears the argument may “fuel the discrediting of scientific expertise”.

Susskind will be giving a public lecture October 17 at UC Davis on String Theory, Physics and the “Megaverse”.

Polchinski avoids the problems associated with the failure of string theory as a unified theory, and promotes in a somewhat overhyped way the idea that string theory explains the RHIC data.

Finally, Smolin makes an offer to string theorists that I feel I should try and match, hoping they will read his book to better understand exactly what he has to say:

If they don’t want to buy it, tell them to get in touch with me and I’ll send them a copy.

One thing Brumfiel gets a bit wrong is that my problem with string theory is not quite what he says “a fear that the field is becoming too abstract and is focusing on aesthetics rather than reality.” The problems I see are rather different, with mathematical abstraction one of the few tools still available to theorists trying to make progress.

The same issue of nature contains an editorial Power and Particles lustily repeating much of the standard hype about string theory, noting that there are problems, but ending with:

Critical-mindedness is integral to all scientific endeavour, but the pursuit of string power deserves undaunted encouragement.

The editorialist definitely does not seem to be of the opinion that alternatives also deserve to be encouraged.

Finally, lots of reviews of Lee Smolin’s book:

Unburdened by proof by George Ellis, also in Nature. Ellis takes the opposite point of view from the Nature editorialist, calling for more research on alternatives to string theory.

A loopy view by Michael Duff, in Nature Physics. Duff is extremely hostile to Smolin’s book, sneering at Smolin and claiming that his book will “leave the reader rooting for strings” (funny, but this doesn’t seem to have been its effect on most reviewers…). Duff agrees that there are problems with string theory, but claims that the problems Smolin correctly identifies are exactly the ones that he himself first identified back in 1987. String theorists like Duff seem torn between claiming that criticisms of string theory are crackpot nonsense, and that they themselves made them first. He goes on to furiously attack various straw men, accusing Smolin of “denying that any progress has been made!” (something I don’t think Smolin does at all), and answering the criticism that string theory makes no predictions despite more than twenty years of effort by discussing how theories that did make predictions have sometimes taken a long time to be confirmed (or remain unconfirmed).

The string theorists were scammed! by Peter Shor on Amazon.

The Trouble With Physics by Sean Carroll at Cosmic Variance. If I can find the time, I may write about some of my problems with this review as a comment over there.

Back in April a paper appeared on the arXiv from string theorist Jacques Distler and collaborators that made a rather outrageously overhyped claim to have found a way to “falsify string theory”. The paper was entitled Falsifying String Theory Through WW Scattering, and was discussed extensively here. After the Wall Street Journal published an article in June about the problems of string theory, Distler wrote them to complain that the article was incorrect, because he and his collaborators had shown that string theory was falsifiable.

I had heard that this paper was going to be refereed, and was wondering whether a referee would really let the authors get away with the outrageous claim of their title. Well, it appears that the answer is no. A new version of the paper is now on the arXiv, with a new title: Falsifying Models of New Physics via WW Scattering. The abstract, which originally claimed that violations of the bounds they described “would falsify string theory” has now been modified to no longer make this claim; the new language is “would falsify generic models of string theory”.

The paper has acquired a new co-author and been extensively rewritten. I’m assuming many of the changes were made to satisfy a referee. Besides changing the misleading, overhyped title, criticisms of earlier work embedded in one reference have been removed, and nine new references to earlier work have been added.

Last week at the KITP, Keith Dienes gave a talk on A Statistical Study of the Heterotic Landscape. He gave a good idea of the state of the art of the investigation of the Landscape, focusing on one special type of models, heterotic models. The results he presented gave statistical distributions for just two very crude aspects of these compactifications, their gauge groups and cosmological constants. These models remain highly unrealistic, since the cosmological constants are of order the Planck scale and the compactifications are not stable.

The models studied have gauge groups of rank 22, and while many of them contain the standard model SU(3)xSU(2)xU(1), they also contain many more gauge group factors, with typically not one, but about seven SU(2) factors. These models, with their instabilities, far too large gauge groups and cosmological constants, are extremely far from anything like the standard model. It’s not at all clear what the point is in enumerating them and studying their statistics, but Dienes describes in detail various problems that arise with the whole concept of generating “random” models of this kind and trying to get sensible statistical distributions. He also looks for correlations between gauge groups and cosmological constants, finding that at small cosmological constant one is somewhat more likely to get many factors in the gauge group (although in his case, both the gauge group and the cosmological constant are very different than in the real world).

Despite the very crude state of these calculations, Dienes reports that a group of 17 prominent string theorists have banded together to form the “String Vacuum Project”, with the goal over the next few years of accumulating a database of 10s of billions of string models, with the hope of finding within this mountain of data about 100 models that have crude features of the standard model. I don’t at all see what the point of this is, but it certainly is a computationally intensive project that could keep many people occupied for a long time. It also appears to be just the beginning, with the longer term goal being to devote the next decades to expanding from 10s of billions farther into the 10^500 or whatever exorbitantly large number is thought to be the number of all string models.

The String Vacuum Project submitted a proposal to the NSF last year, which seems to have been turned down, and they appear to be planning to resubmit the proposal. They have a Wiki, with all sorts of details about the project. Most recent additions to the Wiki are from Bert Schellekens in August, who discusses a proposed “String Vacuum Markup Language” (SVML) format, with links to a web-page that produces data in this format for certain sorts of models. There’s also a European String Vacuum Project web-site.

Fewer and fewer science writers these days are credulous enough to keep promoting string theory, but there still are some around willing to keep writing overhyped stories about how theorists have finally found a way to get some sort of prediction of something observable out of string theory. One of these is Tom Siegfried, who has a new article in Science magazine entitled A Cosmic-Scale Test for String Theory? which reports that “some string theorists now believe they’ve found a way to make superstrings observable.”

Siegfried reports for Science from PASCOS 2006, where he finds two results worth writing articles about. One of these is the recent preliminary neutrino oscillation results from MINOS, which certainly are worth reporting, but the second is the cosmic superstring hype that has been around for nearly three years now, and which I’ve commented on in various places, including here and here. The hype surrounding this topic first got seriously going with a press release from UCSB more than two years ago, in which Polchinski claimed that cosmic superstrings were “potentially visible over the next year or two” at LIGO. Now that this time period is up, the hype has to be modified, and Siegfried informs us that:

LIGO may not be sensitive enough to detect them, but a planned set of three space-based gravitational wave detectors known as LISA would be a good bet.

As is always the case with string theory, there aren’t any real predictions here. The hype is based on the fact that, among the nearly infinitely complicated string theory models people have studied, it is in principle possible to come up with ones in which superstrings created in the early universe would expand to a very large “cosmic” scale and thus be observable. They would show up in various astronomical observations, but no one has yet seen the slightest evidence of such a thing. One can claim that it is logically possible that such things exist, with exactly the right properties to have escaped observation so far, but to be visible to the LISA experiment if it really does manage to get funded and operate sometime in the next decade. While this is logically possible, saying that “it would be a good bet” is pretty absurd; I doubt that any physicist would be willing to put money on this unless given very high odds.

The hype surrounding cosmic superstrings tends to completely confuse the kind of cosmic strings that occur as defects in the Higgs field in some GUT models (which have been studied for about 30 years now) with the kind that are supposed to come from elementary strings. Siegfried’s article includes a graphic purporting to show a “network of enormous ‘superstrings'”. As far as I can tell, this is nonsense, since the same graphic occurs here, in an article from 2000, long before the “cosmic superstrings”, where it is described as showing “cosmic strings form[ed] from a random initial distribution of phases of a hypothetical field called a Higgs field.”

Oh, and the fact that I think this is a pretty sad example of bad science reporting by someone completely taken in by the string theory hype machine has nothing to do with the fact that its author recently wrote an extremely hostile, unfair and inaccurate review of my book…

Update: For an example of the kind of misinformation spread by stories like this, see this blog entry by another science journalist, over at Seed’s ScienceBlogs.

No, I don’t have any idea who will win this year, but the announcement will be a week from today, on Tuesday October 3. After my initial success in Nobel Prize prognostication, I’ve now retired from that game, but encourage others to play.

Well, no, I’m not going to start putting up here the really interesting gossip that people tell me. If I did so they’d stop telling me such things.

The Theoretical Particle Physics Jobs Rumor Mill has moved yet again. First it was hosted at the University of Washington, then the College of William and Mary, now it’s at UC Davis. No idea why it moved this last time, but earlier this year some gossip told me the entertaining story of why it was booted out of Washington. To be honest, I’ve now completely forgotten all the details, so even if I wanted to violate their confidence, I couldn’t.

The new Rumor Mill site confirms previous gossip I had heard that shows UC Santa Barbara having great success in hiring people in mathematical physics. Is Singer has been a regular visitor there in recent years, spending part of the year in Santa Barbara, part at MIT. This year they’ve hired two very good people: Dave Morrison and Sergei Gukov. Morrison has a mathematics background (algebraic geometry), and Gukov was educated as a physicist (a student of Witten’s), but they both do interesting things at the interface of the two subjects.

Also at UCSB, Michael Freedman has moved his Microsoft Research group down from Redmond, and it is now temporarily in residence at the KITP, waiting to move into offices in the building next door when it is finished and will house the California Nanosystems Institute. Freedman is a topologist and Fields medalist, who was hired away from UC San Diego by my ex-grad school roommate Nathan Myhrvold when he was running Microsoft Research. From what I remember, at the time Nathan told me some mildly entertaining gossip about this, but, again, I’ve forgotten the details, so can’t violate his confidence even if I wanted to.

Also on the move is John Horgan’s blog. His Scientific Curmudgeon blog is being shut down, re-opened as a blog hosted by Discover magazine (which has its own blog). The new blog is called Horganism, and he has some advice which I don’t endorse for would-be scientists;

Also, don’t go into particle physics! Especially don’t waste your time on string theory, or loop-space theory, or multi-universe theories, or any of the other pseudo-scientific crap in physics and cosmology that we science journalists love so much.

Seed magazine has some interesting new articles: one by mathematician Jordan Ellenberg about Fields Medalist and MacArthur winner Terry Tao, another by Joshua Roebke about Jim Simons and his Math for America project, the inspiration for which came over a poker game (OK, it was a poker game to raise money for charity).

The Cern Council Strategy Group has put out a briefing book that gives an excellent survey of the prospects for particle physics and particle physics experiments, especially in Europe, during the new few decades. Very much worth reading.

John Baez is encouraging people to join in a campaign to “save New Scientist” from itself, i.e. to get them to stop publishing so much scientific nonsense. This seems to me like a worthwhile goal; maybe if they stop writing articles about crackpots and their “electromagnetic drives”, they’ll also stop promoting bogus over-hyped claims from prominent theorists about cosmology, string theory, etc….

Shing-Tung Yau is fighting back against the New Yorker article “Manifold Destiny”, which was very critical of him, essentially claiming he was trying to steal credit for the proof of the Poincare Conjecture from Perelman. He has hired a lawyer and set-up a web-site. The web-site includes a long letter from his lawyer to the New Yorker, making his case that the article has many inaccuracies. There will be a webcast tomorrow at noon giving his side of this story. Many other blogs and newspapers are discussing this, see here, here, here, here, and here. Unfortunately for Yau, he has strong support from Lubos Motl, who seems a tad obsessed, ranting about how the quality of the New Yorker article:

resembled the style and ethical standards of many jerks in the blogosphere, including a colleague of Sylvia Nasar at Columbia University.

[Note: this has been edited by Lubos, now I’m not a “jerk”, but instead a “despicable writer”]

People who want to engage in bashing of Yau or of his opponents are warned that they should do it elsewhere. Only comment on this here if you have something to say that is substantive and respectful of all parties involved.

Besides Yau’s webcast, tomorrow you can also listen to me on the SETI Radio Network program, broadcast on Discovery Channel Radio. This will also be on their web-site, more info here.

The Harvard Crimson has an article about Nima Arkani-Hamed, who evidently made Popular Science’s “Brilliant 10” list for

his research on the idea that our universe may be only one of many “multiverses” and that additional dimensions may exist.

(many “multiverses”???) Arkani Hamed promotes the anthropic landscape and split supersymmetry as a test for it:

He recently proposed a model for new physics, called split supersymmetry—which theorizes that half of all particles in the universe have partner particles. He said that if the results of the LHC experiment reveal split supersymmetry, “it would be a tremendous push in the direction of a multiverse.”

“Right now a lot of people are on the fence,” about the theory of a multiverse, Arkani-Hamed said. “I think if the LHC sees split super symmetry it’s over.”

Also on the multiverse front, Gibbons and Turok have a new paper out on The Measure Problem in Cosmology. They claim to have a way of determining a measure on the “multiverse”. Only problem is that with their measure, the probability of having inflation work out the way it is supposed to is about e^-180.

Update: Another radio appearance today, on the program This Week in Science.

Update: To view today’s webcast, go to www.premierewebcast.com, get your software working, and enter room 150144. I’ll be skipping this myself, partly because I’ll be in a faculty meeting.

Update: If you want to read a lot of incredibly ill-informed and worthless comments on the Yau story, there’s always Slashdot.