One would think it is only common courtesy to have principal authors review press advisories before they go out. I had no idea there was such potential for miscommunication between science journals and the press. Thanks for posting this.
The issue is important and your views timely. The reason has to do with the, I think, widespread attitude amongst editors of all kinds, who can develop a tendency to know it all, and certainly know more than the guys who did the work. There are some rather famous examples of this who I will not name. The policy of Nature needs changing. Editors are people and can make mistakes, most importantly a mistaken omniscience. Journalists work to deadlines and the compulsion to beat the other guy – reflected in their “exclusive to me/my paper” proclamations.
The public’s right to know, in full, is an ideal worth pursuing. It is even worth more than the work itself (because the work is funded by public money, however disguised) and certainly more than tons of newsprint and/or advisories. thanks for bringing up the topic.
Thanks, Gavin. It’s good to see scientists and others holding Nature accountable for substituting inaccurate PR for accurate science.
You wrote that the principal author “had neither seen nor approved the text of the press advisory.” Now, I’m not advocating changing the subject in this thread any more than you want to discuss the specifics of a stem cell paper, but it simply must be mentioned that this episode calls to mind what Nature famously did to E. A. Foster et al., the geneticists whose scientific report on DNA evidence revolutionized the two-century-old Sally Hemings-Thomas Jefferson paternity debate in 1998.
Nature committed three inaccuracies, the most famous of which was headlining the Foster et al. report “Jefferson Fathered Slave’s Last Child,” when in fact the scientific report showed that _a_ Jefferson, whether or not he was _the_ Jefferson, had fathered the child. (Dr. Foster had not approved that headline, he said at a symposium that I attended.) The claim that none of the other carriers of the Jefferson DNA marker could likely have been the dad may or may not be valid, but in any case that’s a historical, not a scientific, claim, and doesn’t excuse the inaccuracy.
Worse, in an explanatory commentary piece in the same issue, Nature included a summary and a caption, each of which contradicted both the commentary’s authors and the scientific report’s authors by outright falsely reporting that DNA evidence had flat proven that Jefferson was the dad. Whether or not TJ was the dad, the paternity was not flat proven by DNA science. The science only showed that someone carrying a certain DNA marker had been the dad of one of Sally Hemings’s children.
Nature’s job was to convey accurately what science had actually proven, but in those three ways Nature failed. The consequence around the world was that media reports portrayed an absolute DNA paternity proof. This deprived people of the chance to assess, in a nuanced way, what science had and had not actually proven.
That’s bad, and you’re right, Gavin, when you write that “[g]etting things wrong – even subtly – in contentious fields doesn’t help anyone and the slight extra effort to try and prevent mis-communication is well worth it.”
Nature’s editors never forthrightly took responsibility for their miscommunication in the Hemings DNA matter. Thanks for holding them accountable today in your gentle, constructive way.
Comment by Steven T. Corneliussen — 4 Sep 2006 @ 10:30 AM
Press advisories area available to anyone with a fax or email. So, the whackos at CEI and Marshall can then start making the rounds with heads up phone messages tipping reporters to the ‘weakness”, “poor science or junk science” “alarmism”, etc. of the piece in the next issue of Nature.
Why would a scientist want to throw his/her good work into the spin cycle of the AGW denialist washer? I say it is time to drop the press advisories and let everyone read the entire piece collectively. Then, Myron and the gang will have to be more clever in their retorts.
MAY I HAVE THE ENVELOPE, PLEASE actually serves a purpose.
[Response: I disagree. The embargo system and the week's headstart that journalists get in reporting big science stories does work generally to increase the depth of much of the reporting. It takes time to do that and the pressures of time that almost all journalists work under mean that 'news' science would not be even as well treated as it is now should the embargo system per se be removed. My complaint above is only with the lack of input the scientists sometimes have into the pre-publication trailing of their own papers. - gavin]
Comment by John L. McCormick — 4 Sep 2006 @ 10:35 AM
“I would go further and also suggest that the principal authors be given a chance to comment on the advisories before they go out.”
Wow. I really can’t believe this isn’t being done already. All it takes is an e-mail. Just goes to show you common sense and book smarts don’t always go hand in hand.
There’s an underlying problem here, a root cause of which this particular incident is just one manifestation. The embargo system is part of a very successful system by which Nature and Science bless certain papers with a sort of Big Science stamp of approval that often stands in direct contradiction to the substance of the paper.
Reporters are looking for big breakthroughs, but science almost never works that way. Publication in Science or Nature puts an automatic “big breakthrough” stamp on things that might not be – especially given both magazines’ penchant for exciting, frontier science in their decision-making process. To make the media piece of this work, the little one-paragraph press blurb has to accentuate the “breakthroughness” of the paper in question. By their very nature, such frontier breakthrough papers could turn out to be wrong. Nothing wrong with that. That’s the nature of life on the scientific frontier. But the journalist needs a bit of solid ground on which to base the story, and the very publication of the paper in Science or Nature implies that.
The Bryden paper on the thermohaline circulation in Nature last fall is a great example (see the RC discussion if you’ve forgotten – this was the paper suggesting a slowdown in the thermohaline circulation).
Re #5 (and others surprised that authors don’t get to give feedback on the press advisory): My guess is that Nature would say that if given the opportunity, authors tend to nitpick and then they have to have back-and-forth on the wording and that this really slows things down and creates a lot of extra work.
Mind you, I am not justifying them not getting authors’ approval on the press advisory but merely expressing why it might create a little more work than people here imagine and why this might not be quite as much of a no-brainer as it appears to be. Perhaps Nature could send the press advisory to authors with the caveat saying something like “Please do not try to rewrite our advisory as you would have written it but merely stick to correcting serious errors of fact or interpretation. If you want something re-worded, please provide a full explanation of why the current wording is seriously in error.”
Why don’t the editors simply ask the authors to write a “so many-word” press advisory for their papers when they send authors acceptance letters? This way it would be less work for the editors. They could give specific instructions to authors about pitching the advisory for journalists and a general audience. And the editor would still have control over what will be said in the press advisory because they would have to edit it. It could be like the paper will not be published until the authors provide an acceptable press advisory which would motivate authors to be timely with it.
[Response: I think the dictum 'from each according to their talents' might be applicable here. Us scientists are good (in general) at many things - crafting pithy, accurate, attractive, and tempting paragraphs describing their work for non-scientists is not (in general) one of them. Editors are usually much better at this in my experience - only occasionally do they need a tweak in the accuracy department. - gavin]
Ok, I see your point, Gavin. Then how about provide authors with the press advisory with the letters of acceptance or even later in the process with the copyright transfer forms perhaps. It could be one of the things authors have to sign off on before publication and like number #8 saya with clear instructions about only checking for factual and interpretive accuracy. So the authors can’t edit the advisory but they could offer suggestions if something is misrepresented.. That can’t be to time intensive for the editors.
Interesting, I didn’t know they sent out press releases a week in advance. I’ve published in both Nature and Science and didn’t receive any communication from the journals regarding a press release in advance. Then again, what is there to prevent authors from drafting their own press releases with university PR departments in advance of publication? I think scientists and universities should take a more active role and managing their own press.
[Response: Absolutely. But the first thing that the journalists see are the press advisories. Where the authors and institutions do their own releases, these generally dominate the coverage. -gavin]
Would today’s stories, like this one as found at sciam.com today be based on Nature press releases?
Anyone got an original, or more info on this one?
By Patricia Reaney
LONDON (Reuters) – Greenland’s massive ice sheet is melting much quicker than scientists had estimated and the pace has accelerated lately, according to research published on Wednesday.
…. An analysis of satellite observations shows the rate of ice loss rose 250 percent between the periods April 2002 to April 2004 and May 2004 to April 2006, most of it in southern Greenland…..
The findings, based on data from the Gravity Recovery and Climate Experiment (GRACE) satellite, and published in Nature magazine, are consistent with earlier results showing increased melting due to rising temperatures blamed on global warming.
… their analysis is very recent, up to April 2006, and shows the accelerated rate of loss is almost entirely in southern Greenland.
“… a very strong acceleration, which means things are changing,” said Velicogna.
—– end snippet——
Nature does not normally issue full press releases about individual articles. The ‘press advisory’ snippets that you refer to are one or two paragraphs summarising the research. We (I speak as a science journalist) receive them four or five days in advance, and yes, this does help us do more thorough reporting.
Press releases written by research institutes (most have a press office these days) can appear any time, depending on the press officer. Very often, they are posted on Eurekalert (www.eurekalert.org), which is run by AAAS and to which most US and UK science reporters subscribe.
Journalists would also prefer that press snippets are run by the leading author before they are issued to us — it would make our job easier. But for this, scientists need to be more aware of the press ‘mechanisms’ that accompany their publication. Journals need to ask that they be available to review a press statement very quickly a few days before publication. And scientists have to be willing to do this.
I have on a few occasions called scientists who had just or were about to have a paper appear in a leading journal only to find that they were on leave or unreachable at a field site. There can be two reasons for this: either the journal has not informed them of the press process or the scientist does not believe this aspect of publishing is be sufficiently important to care. (In fairness, though, this situation is becoming less and less frequent.)
Mind? THANK YOU. Much appreciated, please stay around and say more from your perspective.
Reading that press release you link to makes me think this story’s being under-reported:
“”These findings are consistent with the most recent independent measurements of Greenland’s mass done by other techniques like satellite radar interferometry, but in this case they provide a direct measure of ice-mass changes,” said Geology Professor Clark Wilson, a co-author on the latest Science article who helped analyze the estimates for Greenland. Wilson chairs the Department of Geological Sciences at the university and holds the Wallace E. Pratt Professorship in Geophysics.
Within the subpolar zone that includes Greenland, the rapid rise in meltwater along its eastern coast could add to other warming-related factors believed to be weakening the counterclockwise flow of the North Atlantic Current. ….
“If enough fresh water enters the Norwegian Current,” Tapley said, “and you interrupt return flow, then there could be climate effects in Europe.”
“This melting process may be approaching a point where it won’t be centuries before Greenland’s ice melts, but a much shorter time-frame,” Tapley said, noting that it isn’t possible to tell how much sooner this will be.
==== end quote =====
Following that up, here’s a very similar release dated 9/20, minus that last bit about Greenland melting much sooner than expected, seems to be the same study discussed but it’s out of U. Colorado. Odd.
Re 14 – 17 Hank there is a difference. The first paper in Science was upto 2005, while the later paper in Nature is upto 2006.
One point that no one seems to make is that if the Greenland ice sheet does melt and sea level rises 7 m it will raise the Antarctic ice shelves by 7 m too. How much would they have to be raised before they break off. Less than 7 m is my guess :-(
I think the only reason stem cells therapies are not being pursued by the United States Governemnt is becasue the Presidents ties to the bio tech industry and all these companies want to patent the therapies instead of them being given out.
Good point in 18, Alastair — each incremental increase in sea level is going to force sea water in between the grounded ice and whatever it rests on, too. I’d think that it’d be melting at that interface. I suppose GRACE can detect that, it apparently can detect different temperature ocean currents or layers, I gather. Must be a lot of data, with 15 orbits per day.
It seems to me that sea water intruding under a grounded ice shelf would appear to the GRACE satellite to be a thickening of the shelf. Because GRACE measures changes in the mass beneath its orbit, it can’t distinguish between mass added above the ice shelf (snow and ice) and mass intruding below the ice shelf (seawater). There exists the potential for the ice shelf to appear stable or growing . . . right up until it breaks up. Shades of the Larsen B breakup!
Would sonar be able to detect if a liquid water layer exists between the ice sheet and bedrock, and if so, how deep it is?
Also, sea water intrusion should cause a discrepancy between altitude readings and gravity readings, since sea water is denser than ice and air bubbles. Would this signal be strong enough to sort out of the noise?
Ice-penetrating radar may be able to image the grounding line, that is, the transition zone where a grounded ice sheet becomes a floating ice shelf. This type of radar has been very useful in detecting and measuring lakes beneath the Antarctic ice (Lake Vostok is the largest and best known of these lakes). If the global rise in sea level is lifting the ice shelves and allowing seawater to intrude into new areas then the grounding line should be shifting inland in a measurable fashion.