Some of the world’s finest physicists and cosmologists have in recent weeks been pouring scorn on the now infamous OPERA result. (If you’ve just been released from one of those Mars simulation missions, such as Mars500, then I guess it’s possible that you might have missed what has the potential to be the biggest physics result in a century: the report by the OPERA collaboration that muon neutrinos produced by CERN travelled ever-so-slightly faster than light while on their way to detectors at Gran Sasso.) I’m sure that those scientists, many of whom I admire tremendously, are right: those neutrinos are surely not travelling faster than light. It wasn’t as if the neutrinos acted like resublimated thiotimoline, somehow arriving at the OPERA detectors before they were produced. The OPERA team were making tremendously difficult measurements, and at this point it’s safer to assume that their finding is the result of some unknown source of error in the experiment. But there’s one point on which I think those eminent critics of OPERA have it wrong.
The criticism is that the OPERA team contacted the media and called a press conference before they published their results in a peer-reviewed paper: irresponsible behaviour, clearly, particularly where such a controversial result is involved. Thing is, the OPERA researchers didn’t announce their results at a press conference: they announced them at a CERN seminar. And they didn’t draft a press release: they submitted a technical preprint to arXiv. Surely they did everything that responsible scientists should do?
Once, not many years ago, you could put a preprint on arXiv and you knew you’d be reaching an audience of physicists. We now live in a world of blogs (well, you’re reading this one aren’t you?) and Twitter. Put a preprint on arXiv that says in effect “Einstein was wrong” and you may as well shout it out loud while standing naked at Speaker’s Corner. Perhaps unfortunately for OPERA, in the modern world of social media there’s no way that the original seminar could go unnoticed; the press conferences that followed were inevitable – and then so was the criticism that the collaboration hadn’t followed proper processes.
If there’s a criticism to be made of OPERA it is, I believe, that they hadn’t ruled out all sources of systematic error before giving that initial CERN seminar. Indeed, that’s probably why ten senior members of the collaboration decided not to sign the arXiv submission. One obvious concern with the experiment, which many physicists voiced immediately, is that CERN was sending long neutrino pulses (about 10 microseconds long) to Gran Sasso; the effect they were observing, though, involved a shift that was a tiny fraction of that pulse length (the shift was about 60 nanoseconds). For their analysis to work, the collaboration needed to know the shape of the neutrino pulse quite precisely; but they were only able to infer the neutrino pulse shape. (The neutrinos come from protons smashing into a target; OPERA infer the neutrino pulse shape from the initial proton pulse shape.) Get that inference just a little bit wrong and they would end up seeing things that just aren’t there.
Fortunately, there’s a really simple way to get round this difficulty: repeat the experiment, but send a series of short neutrino pulses separated by large gaps. That way you don’t need to know the neutrino pulse shape: each pulse from CERN is unambiguously linked to the OPERA detector.
The OPERA collaboration has now run precisely this experiment. They asked CERN to generate proton pulses lasting just 3 nanoseconds, and recorded 20 neutrino events. And the result? Well, again the neutrinos reached Gran Sasso about 60 nanoseconds before light itself could have reached there. The anomaly remains.
So when will the fat lady sing at OPERA? When will we know what systematic error is to blame for this bizarre result? (And for what it’s worth I think it will turn out to be a systematic, probably to do the use of GPS in the experiment.) Well, it’s clear that independent checks are required. The first project to be in a position to do those checks is likely to be MINOS at Fermilab. We might get results from MINOS some time in 2012. If MINOS replicates the OPERA result… well, then we’ll be living in interesting times.