Skip to main content
Publishing

Publishing

The brave new-media world

10 Nov 2011

When physicists in Italy claimed to have seen hints of faster-than-light neutrinos, it initiated a media frenzy as well as criticism of publicity-seeking. But as Chad Orzel argues, the scientists behaved just as they should

Out in the open

On 23 September scientists at the Gran Sasso underground laboratory in Italy announced a discovery that could potentially revolutionize physics. The OPERA collaboration based at the lab found evidence that muon neutrinos produced by the CERN particle-physics lab near Geneva move slightly faster than the speed of light while on their way to Gran Sasso, where they are detected. If confirmed, the finding would put into question Einstein’s special theory of relativity, which forbids superluminal travel.

The result quickly turned into one of the most covered physics stories of the year, with numerous articles in magazines, newspapers and on television asking whether “Einstein was wrong”. Just as quickly came numerous physicists denouncing the media frenzy, with Lawrence Krauss from Arizona State University and Cambridge University cosmologist Martin Rees both calling the coverage “an embarrassment”.

“A press conference on a result, which is extremely unlikely to be correct, before the paper has been refereed, is very unfortunate – for CERN and for science,” Krauss told Scientific American. “Once it is shown to be wrong, everyone loses credibility.”

A closer look, however, suggests that the OPERA researchers behaved exactly as scientists should. They did not write a press release, but a technical preprint on the arXiv preprint server; and they did not schedule a press conference, but a seminar at CERN. While some of the media coverage has been regrettable, OPERA scientists are not the architects of overhype, but the victims of a radically changed media landscape.

Ahead of the game

Every sub-field of physics has its stories about how responsible physicists deal with anomalous results. In my own speciality of atomic, molecular and optical physics, for example, Bill Phillips from the National Institute of Standards and Technology (NIST) discovered in the mid-1980s that laser-cooled sodium atoms had a temperature of 40 µK – well below the theoretical limit of 240 µK. Confronted with the extremely odd situation of an experiment working better than expected, Phillips’ team re-checked its results, in the process inventing five new ways of measuring the temperature. When all of their measurements gave the same results, they quietly contacted other researchers in this relatively small community and also gave seminars at other labs. Phillips asked these labs to check their own atoms’ temperature and only after receiving independent confirmation did the NIST group publish the result (Phys. Rev. Lett. 61 169).

The discovery spurred other labs to work on explaining the lower temperatures and groups led by Steve Chu at Stanford University and Claude Cohen-Tannoudji at the Ecole Normale Supérieure in Paris soon worked out the correct theory. A decade later Phillips, Chu and Cohen-Tannoudji shared the 1997 Nobel Prize for Physics for their work on laser cooling. This case is often cited as an example of how to handle surprising results: first re-check all the measurements; then seek independent confirmation; and only after confirming your results, schedule a press conference.

Yet this is essentially what the OPERA collaboration did. Their paper on arXiv (1109.4897) shows that they considered most of the obvious sources of error in their results and re-checked the key elements by, for example, having national standards labs, such as the Swiss Metrology Institute and the PTB in Germany, verify the synchronization of their clocks and the distance between the neutrino source and the detector. There is some debate about whether they exhausted all possible checks, and around four senior members of the 160-strong collaboration removed their names from the paper as a result (see p12, print edition only). But they did also have some support from an earlier measurement in 2007 when the US-based MINOS experiment reported a similar anomaly in the apparent speed of its neutrinos, though with larger experimental uncertainty.

With the analysis showing an anomalously high speed, the appropriate next step was to present the result to other physicists to check the results. In this case that meant posting the preprint on arXiv and scheduling a seminar at CERN to present the results to leading particle physicists. Thanks to blogs, Twitter and other forms of social media, however, posting a preprint and scheduling a seminar are more or less equivalent to calling a press conference. Indeed, science journalists routinely monitor social media and arXiv for stories. Just 15 years ago, posting a paper on arXiv was a quiet way to disseminate results to other physicists; today, it is as good as e-mailing it to every science reporter.

The new-media landscape

Research is increasingly being done in the open. Some scientists do this deliberately, by posting all their data on freely available websites; but for others, the new openness is not by choice. And while physicists pioneered open-access science with arXiv, in many other ways we have been slow to adapt. Indeed, the recent brouhaha is not the first time physicists and social media have collided, with, for example, particle physicists Tommaso Dorigo and John Conway sparking controversy in 2007 by discussing preliminary data from Fermilab’s Tevatron collider on their blogs. Rumours of preliminary data spreading through social media have also led to some false alarms, such as the inconclusive findings of the CDMS dark-matter search in late 2009 being blown up into major events.

The problem of exciting results being released early will not go away and if anything is only likely to get worse. Physicists, therefore, need to adapt to the brave new-media world, in which it is nearly impossible to keep exciting results completely under wraps. Physicists are justly proud of having created the World Wide Web; now we have to get used to doing science in full view of the Internet.

Related events

Copyright © 2024 by IOP Publishing Ltd and individual contributors