UCal Irvine’s ‘fifth force’ farce

Michael Moyer just concluded a rant on Twitter (at the time of writing this) about how a press release on a recent theoretical physics result developed at the University of California, Irvine, had muddled up coverage on an important experimental particle physics result. I was going to write about this in detail for The Wire but my piece eventually took a different route, so I’m going to put some of my thoughts down on the UCI fuck-up here.

Let’s begin with some background: In April 2015, a team of nuclear physicists from the Hungarian Academy of Sciences (Atomki) announced that they had found an anomalous decay mode of an unstable beryllium-8 isotope. They contended in their paper, eventually published in Physical Review Letters in January 2016, that the finding had no explanation in nuclear physics. A team of American physicists – from the University of California, Irvine, and the University of Kentucky, Lexington – picked up on this paper and tried to draw up a theory that would (a) explain this anomaly even as it (b) would be a derivative of existing theoretical knowledge (as is the work of most theoretical physics operating at the edge). There are many ways to do this: the UCI-UKL conclusion was a theory that suggested the presence of a new kind of boson, hitherto undiscovered, which mediated the beryllium-8 decay to give rise to the anomalous result observed at Atomki.

Now, the foreground: A UCI press release announcing the development of the theory by its scientists had a headline that said the Atomki anomalous result had been “confirmed” at UCI. This kicked off a flurry of pieces in the media about how a ‘fifth force’ of nature had been found (which is what the discovery of a new boson would imply), that all of physics had been overturned, etc. But the press release’s claim was clearly stupid. It was published no more than a week after the particle physics community found out that the December 2015 digamma bump at the LHC was shown to be a glitch in the data, when the community was making peace with the fact that no observation was final until it had been confirmed with the necessary rigour even if physicists had come up with over 500 theoretical explanations for it. The release was also stupid because it blatantly defied (somewhat) common sense: how could a theoretical model built to fit the experimental data “confirm” the experimental data itself?

There’s even a paragraph in there that makes it sound like the particle’s been found! (My comments are in square brackets and all emphasis has been added:)

The UCI work demonstrates [misleading use] that instead of being a dark photon, the particle may be a “protophobic X boson.” While the normal electric force acts on electrons and protons, this newfound [the thing hasn’t been found!] boson [a boson is simply one interpretation of the experimental finding] interacts only with electrons and neutrons – and at an extremely limited range. Analysis co-author Timothy Tait, professor of physics & astronomy, said, “There’s no other boson that we’ve observed that has this same characteristic. [Does this mean UCI has actually observed this particular boson?] Sometimes we also just call it the ‘X boson,’ where ‘X’ means unknown.”

Moyer says in one of his tweets that PR machines will always try to hype results, outcomes, etc. – this is true, and journalists who don’t cut through this hype often end up writing flattering articles devoid of criticism (effectively missing the point about their profession, so to speak). However, as far as I’m concerned, what the UCI PR has done is not build hype as much as grossly mislead journalists, and I blame the machine in this case more than the journalists who wrote the “fifth force found” headlines. Journalism is already facing a credibility crisis in many parts of the world without having to look out for misguided press releases from universities of the calibre of UCI. Yes, such easily disturbed qualities are also often trusted by journalists, or anyone else, because we trust institutional authorities to take such qualities seriously themselves.

(Another such quality is ‘reputation’. Nicholas Dirks just quit because his actions had messed with the reputation of UCal Berkeley.)

This is a problem exacerbated by the fact that journalism also has a hard time producing – and subsequently selling – articles about particle physics. Everyone understands that the high-energy physics (HEP) community is significantly invested in maintaining a positive perception of their field, one that encourages governments to fund the construction of mammoth detectors and colliders. One important way to maintain this perception is to push for favourable coverage in the mainstream media of HEP research and keep the people – the principal proxy for government support – thinking about HEP activities for the right reasons. The media, in turn, can’t always commission pieces on all topics nor can it manufacture the real estate even if it has the perfect stories; every piece has to fight it out. And in crunch times, science stories are the first to get the axe; many mainstream Indian publications don’t even bother with a proper science section.*

If, in this context, a journalist buys into a UCI press release about some kind of ‘confirmation’ of a fifth force, and which is subsequently found to be simply false, an editor wouldn’t be faced with a tough choice whatsoever about which section she has to axe.

What happens next? We wait for experimental physicists try to replicate the Atomki anomaly in experiments around the world. If nothing else, this must happen because the Atomki team has published claims of having discovered a new particle at least twice before – in 2008 and 2012 – both at a significance upwards of 3 sigma (i.e., the chances of the results being a fluke being 1 in 200,000). This is a statistical threshold accepted by the particle physics community and which signifies the point at which a piece of data becomes equivalent to being evidence. However, the problem with the Atomki results is that both papers announcing the discoveries were later retracted by the scientists, casting all their claims of statistical validity in doubt. The April 2015 result was obtained with a claimed significance of 6.8 sigma.

*Even The Hindu’s science page that used to appear every Thursday in the main newspaper was shunted last year to appear every Monday in one of its supplements. It never carried ads.