The PNAS problem: When papers aren’t available when the embargo lifts

The Proceedings of the National Academy of Sciences (PNAS) sends out embargoed press releases every Wednesday for a selection of papers that will be published the following week. The embargoes lift the following Monday at 3 p.m. Eastern. That’s a fairly typical embargo setup.

What’s not as typical is that some of the other papers that will be published in the same week — whose embargo also lifts at 3 p.m. on Monday — are available to reporters the previous Wednesday, but not available online when the embargo lifts. They become available sometime during the week, often as late as Friday. That means that when stories based on those studies come out, you’d search in vain if you wanted to find the original papers.

Strictly speaking, this isn’t an embargo issue per se. It’s more about what happens right after embargo. And it has a lot of science bloggers who tend to write about those papers pretty irritated. (Their readers, too.) Science blogger Ed Yong calls this “post-embargo publication window” a “complete anathema to the internet world.”

People expect to be able to tumble down the rabbit-hole of links to find original sources and check them out for themselves, if they are so inclined. These outmoded policies mean that the rabbit hole ends in 404 purgatory.

The problem is not unique to PNAS, he notes:

For some journals, this time is negligible – Nature and Science, for example, reliably have their papers up within minutes or hours of the embargo lift. For others, it can be much longer. PNAS is the most obvious example – I’ve waited for up to two weeks before the paper actually went online after it made the news. The record so far is several months for a Journal of Zoology paper.

Still, PNAS’‘ volume and frequency of papers puts them in the spotlight, so I contacted them for their rationale. (Here’s their embargo policy.) Jonathan Lifland, the journal’s media and communications manager, told me by email that “the answer is perhaps more simple than you might imagine.” He went on:

We publish online daily as the articles are finished by our editors and reviewed/approved by the authors, which happens at random intervals during the week. We arrange for our designated press-interest articles, which includes the articles listed on the weekly PNAS “Tipsheet,” to publish on Monday so that the media and public are able to access those papers that we hope that they will find interesting. However, the sheer number of articles that we publish makes it impossible for all of the research to be available at one time.

(There are some exceptions to the Tipsheet policy. For example, a study on oxytocin and autism, whose embargo lifted on February 15, wasn’t posted online until the 16th because of the President’s Day holiday the 15th.)

I suppose journalists could refuse to publish stories on papers until they’re available online. But I think that’s highly impractical. If an embargo breaks, one or more of your competitors is going to post something. Waiting until you can link to a paper — or until a reader can find the paper, since it’s not as easy as it should be to add URLs to stories, particularly on a wire — seems like the best service for readers. (Some of my journalist colleagues might argue that very few readers will bother, and they may be right. That doesn’t mean we shouldn’t try to do it.) But it’s also a good way to lose traffic and clients to competitors.

Ed and others also suggest that journals could use document object identifiers (DOIs) better. DOIs are now assigned to every published paper, and basically provide a formula for finding a URL for a given paper forever. If journals assigned those ahead of time — which certainly seems possible — they could include them in press releases. But the paper would still have to be available at the time the embargo lifted, which seems to be the sticking point.

The complaints I’ve seen suggest that some writers would prefer if PNAS just waited until they knew the studies were available to be put online before sending out the releases. Get everything set up on the site, leave it all in draft, and send out an embargoed press release. Resource constraints might make this difficult, as Jonathan suggests. But it seems like the only realistic way to solve the problem.

Thoughts?