I was concerned that my “this week’s hype” headlines might be less than accurate, counting only 3 separate over-hyped string theory stories during the past month rather than four. Turns out that I missed one (although a commenter here didn’t), involving yet another university press release based on a PRL-published paper about cosmic strings. To be fair to the authors, the press release and paper don’t contain that much hype, nothing about how they are “testing string theory”. What they do is fit the CMB data using an additional parameter they call f10, which has to do with the fractional contribution of cosmic strings to the temperature power spectrum at multipole l=10. They claim to get a slightly better fit to the data with a non-zero version of this parameter and power-law tilt ns=1, versus the usual fit with gives a ns less than one. When they also take into account non-CMB data, the effect goes away.
This isn’t really convincing of anything, so it’s unclear why it deserves a press release. According to the New Scientist story on this, which is pretty reasonable and hype-free, the chief scientist for WMAP, Charles Bennett thinks it’s a statistical fluke:
Calling it a detection is odd… I’d be very surprised if cosmologists were excited about this at this stage.
For other press stories about this, featuring misleading headlines, see String Theory Gets A Boost at physorg.com, and String Theory slightly preferred… or at least, not disfavored! at Canada Free Press, where the author does note:
To listen to people speak about string theory is a lesson in ambiguity. No one is willing to commit to a solid opinion, on either side of the coin, and they dance upon the fence as if they were auditioning for a Garfield strip.