Science: Scramble before the facts

You were not supposed to know until next Friday that scientists at Nasa thought they had found traces of past life on Mars. The announcement was meant to happen in a blaze of carefully orchestrated publicity, to coincide with the publication in the weekly journal Science of the researchers' carefully-argued paper, "Search for Past Life on Mars: Possible Relic Biogenic Activity in Martian Meteorite ALH84001".

So why have we all heard about it already? Because of that modern phenomenon in science publishing - the irresistible thirst for news. Other examples are easy to come by: the discovery last October of the first planet orbiting a Sun-like star in the Milky Way; the recent disclosure that BSE, or mad cow disease, may be passed from mother to calf; the British government's announcement in March that a dozen recent cases of Creutzfeldt-Jakob Disease (CJD) may have been caused by BSE. In the latter two cases, a paper for a scientific journal was being prepared when the announcement was made, but neither had yet been accepted for publication.

What's wrong with that? To the general public, nothing. Certainly, the news about life on Mars has spiced up what might otherwise have been a dull week. But to many scientists, such "pre-announcements" can leave them shifting uncomfortably, especially when eager journalists call them to ask for a quote. The worry is that it undermines the scientific process, which is meant to ensure that only carefully researched work is published.

Scientists dislike commenting on work that has not been published and might still be undergoing the "peer-review" process (in which scientists in the same field read the paper to decide whether the research has been properly carried out). Certainly, in the case of BSE passing from cow to calf, statisticians have been arguing furiously over the Government's interpretation of the figures (that it shows the chances of transmission are 1 per cent). Many disagree strongly. The final paper may offer a quite different number.

The case of "life on Mars" is slightly different: the paper had been peer-reviewed and accepted. But the way the news emerged - sparked off by a single paragraph headed "Meteorite Find Incites Speculation on Mars Life" in the technical publication Space News - was completely different, prompting a scramble for Mars experts, any experts, by the world's media.

"We don't necessarily like it when this happens," says Diane Dondershine, in charge of publicity for Science. "The process we have set up" - by which science journalists in a large number of media organisations are told, under embargo, what papers will be in Science the following Friday - "means reporters can do background research and all get their stories at the same time. I don't think it's the best thing that can happen if everyone's scrambling to get a story."

Certainly, there have been times when pre-announcement has led to wild speculation which has later proved to be unfounded. Do the words "cold fusion" ring a bell? In March 1989, Stanley Pons and Martin Fleischmann called a press conference in which they described benchtop experiments that were interpreted as showing fusion in a test-tube of deuterium - a "heavy" isotope of hydrogen - to create helium.

There was a paper to go with the announcement, but it had not yet been published (in the Journal of Electrochemical Chemistry). Fleischmann's work plan didn't intend publication until September of that year. After the press conference, which led the media to begin writing excitedly of a pollution-free power source which could be powered by seawater (which those with longer memories will recall happened in the 1950s, with the first "hot fusion" experiments), other nuclear scientists were quick to rubbish the claims.

As was pointed out, if there really had been fusion going on in those test-tubes, both experimenters would have been killed by the neutron shower released in the process. If there is one thing that scientists hate, it is thinking they have been duped. The same goes for journalists. Consequently, "cold fusion" has become the science pursuit that dare not speak its name (though it is still active, as a brief perusal of a few pages on the World Wide Web demonstrates).

"When results are released ahead of the peer-review process, it's much more dangerous both to the journal and to the whole scientific process," says Laura Garwin, physical sciences editor of Nature. "In the case of the 'first planet', the result was not released to the media by the researchers, but it got out. The trouble was that their paper changed between the announcement and publication."

Of course, there is a way to ensure that everybody gets a simultaneous view of the paper: publish it on the Internet. Then everyone with a computer and telephone line can see it and judge for themselves. Let the peer review occur after publication. Paul Ginsberg, the physicist, has been quoted as saying he doesn't need peer review to tell him if a paper is good or bad, and that he sees both in peer-reviewed journals.

But although many US and British scientists do have high-speed access to the Internet, the indications are that it is not going to take the place of Nature or Science in a hurry. "The Internet has existed for a long time for physicists and astronomers," says Ms Garwin. "But not everybody is like Ginsberg. In a way, the Internet makes peer-reviewed journals much more valuable: there's so much information washing around, but we make it easier to deal with. Reading Nature means you only have to deal with a tenth, no, a hundredth, of what's out there. What we offer is quality control, now that the Internet has made 'publishing' so easy."

Dondershine echoes those views, though she notes that the global network does have its uses in a publishing crisis like last week's. "In this case it was very helpful. We were able to get the it up on our Web page, including all the visuals and the information, on Tuesday night. It made the whole process much faster than it would have been 10 years ago."