Several people have pointed out to me the latest press release hyping the supposed testability of an extremely speculative theoretical idea, which then gets promoted to a “scientists finally find way to test string theory” story, and spread throughout the popular press.
This week’s hype example comes from Duke University and is entitled Scientists Predict How to Detect a Fourth Dimension of Space. It deals with a recent paper by Keeton and Petters, one in an interesting series of papers about using gravitational lensing to test GR. This latest paper deals with possible lensing effects of primordial black holes in braneworld models. The hype isn’t really in the paper itself, but in the press release, where Petters says “If braneworld black holes form even 1 percent of the dark matter in our part of the galaxy — a cautious assumption — there should be several thousand braneworld black holes in our solar system.” Braneworld scenarios can have any energy scale one wants, and the only thing one knows about this is that it can’t be below a TeV or so, because otherwise we’d have some evidence from accelerators for these scenarios, and we don’t have a shred of such evidence. I just don’t see any justification for calling the idea that 1% of dark matter is made up of these black holes a “cautious assumption” about what a braneworld scenario would “predict”, or for claiming that they have a testable “prediction” for what the GLAST satellite will see, in any conventional scientific use of the word “prediction”.
As usual, the hype level increases as the story is reworked into popular science stories elsewhere. For instance, at Ars Technica it is the inspiration for an article called String theory makes prediction – pig grows wings. The writer begins by giving a completely incorrect explanation of why string theory has made no experimentally verifiable predictions to date:
because the governing equations which work so well at very small scales (and very high energies) become impossible to solve when applied at lower energy or larger scales. Thus theorists must make approximations, which then have another layer of approximation applied before any measurable numbers fall out. At this point everything falls apart because that second layer of approximation is governed by existing experimental results, which means that no new predictions are made.
this repeats the usual misleading claim of string theorists that they have “equations which work so well at very small scales”, and the only problem is that it is hard to extract physics at long distance scales from these equations. I have no idea what this “second layer of approximation” is, best guess is that it is the “approximation” of assuming that string theory gives you the standard model at low energies, an “approximation” that does kind of make it hard to extract predictions that disagree with the standard model.
The writer also refers to another bogus “prediction of string theory” he attributes to Ulf Danielsson, and concludes that “now we have some real testable predictions from a theory of gravity derived (not in the mathematical sense) from string theory.” The parenthetical remark at least gives some indication that something very fishy is going on, and he does end the piece by pointing out that:
A word of caution should be attached at this point. Braneworld gravity is one of a number of string theory derived candidates, so if braneworld fails don’t expect it to take string theory down with it.
Update: As usual, picked up by New Scientist.