Have Physicists Just Disproved String Theory?

No. The answer is no. But you’d be forgiven for thinking otherwise, because string theory has had a really rough couple of years. This is because physicists had two promising opportunities to prove the accuracy of a simple string theory model, and neither opportunity panned out.

The first serious opportunity came when CERN’s Large Hadron Collider went live. As Columbia University’s Brian Greene explains in his 2008 introduction to string theory, the LHC had the potential to prove string theory correct:

[embed]https://www.youtube.com/watch?v=YtdE662eY_M[/embed]

Now it’s six years later, and we know that isn’t going to happen, at least not the way Greene and others had hoped. The LHC has not detected what we would expect it to detect if a simple string theory model were accurate, which probably means that either string theory isn’t accurate, or the particular version of string theory we accept needs to be more complex. A lot of physicists thought it did anyway; Greene himself identifies a wide range of string theory models, and most of them are not affected in any way by the lack of LHC data. What the lack of proof really does is help refine string theory to exclude some of the simpler models.

Likewise, the German study that seemed to confirm Lorentz invariance earlier this week—when a small number of string theory models predicted that Lorentz invariance could be violated—doesn’t really affect the overall validity of string theory. A number of laypersons are likely to believe that string theory has been disproven by this study, and it hasn't. (The American Physical Society’s statement that "certain quantum gravity theories, such as string theory, predict [Lorentz invariance] violation at a very small level” doesn’t help matters; most string theory models do not predict measurable violation of Lorentz invariance.) As with the LHC data, string theory has been refined, not disproven.

Here’s a good point of comparison: let’s say you were somehow able to comprehensively search Yellowstone Park and prove Bigfoot isn’t there. That doesn’t mean Bigfoot isn’t real; it just means he isn’t in Yellowstone Park, so you know you can stop looking for him there. That’s fine. But if you were able to comprehensively search Loch Ness and prove Nessie isn’t there, that would be a much bigger deal because you’d have categorically proven that the Loch Ness monster doesn’t exist, period, full stop; if you found Nessie anywhere else, it wouldn’t be Nessie.

Some string theory skeptics have reacted as if the LHC and Lorentz invariance data have created a no-Nessie scenario when they’ve really created more of a no-Bigfoot-in-Yellowstone scenario. A few string theory models have been discredited by recent studies, but the vast majority of string theory models are no more or less likely to be accurate than they were ten years ago. And since it’s impossible for more than one model of string theory to be accurate anyway, that isn’t really a gamechanger.