New Scientist had the good sense to pass on last week’s hype about string theory testability, but is responsible for this week’s hype on the subject, with an article entitled String theory may predict our universe after all. It’s unclear why the article is appearing now, since it is based on a six-month-old preprint from a group at Oxford entitled Triadophilia: A Special Corner of the Landscape.
The authors basically just point out that there are very few known Calabi-Yau manifolds with small Hodge numbers, which thus have a small enough Euler characteristic to give just three generations. They speculate that some unknown dynamical vacuum selection mechanism favors these particular manifolds. In their paper they look only at the topology of the manifolds, so the only “prediction” about our universe is that the number of generations will be small, and this “prediction” is based on assuming an unknown dynamics that favors small numbers of generations.
There has been a huge industry since the late 1980s devoted to trying to extract physics out of the sorts of Calabi-Yaus studied by the Oxford group. This hasn’t gotten very far, with rather elaborate mathematical constructions being used to try and get the quantum numbers of the standard model particles to come out right. One problem with this is that one is not even sure that this is what one wants, since maybe the LHC will find more particles. The groups pursuing this strategy don’t seem to have taken much interest in the Candelas et. al paper, since SPIRES shows that no one has cited it during the last six months.
It looks like 2008 is not going to show any slackening of the promotion by string theorists of bogus “Despite what the critics say, string theory really is predictive!” stories to the press. This one contains quotes from Polchinski that the paper is “neat” and “Maybe it gives us a clue”, and from Strominger that it is “beautiful”. Strominger also minimizes the fact that the Landscape is a problem for string theory, saying:
I don’t think it is incumbent upon string theory to solve the problem of the landscape… If we can’t make the landscape go away, it doesn’t mean that string theory is wrong. It just means it is not a complete solution to all our problems.
Michael Duff says the paper makes “some mathematically sound and interesting observations”, but does note that it doesn’t explain what selects small Hodge numbers, which is about the only slight amount of non-hype that makes it into the article.
Update: As a commenter here points out, the New Year also brings new progress on the scientific investigation of the landscape/multiverse, with a preprint from Don Page about how God loves all universes, not just ours.