Secrecy in science

Scientists generally advocate for openness. Full disclosure of methods is vital to peer review and to reproducibility or even evaluation of experimental results. Scientists are also pushing hard for a new publishing system which doesn't hide research behind copyright walls. The community of science is largely an open book, encouraging collaboration, review, and public discussion of new findings. So it's understandable that Kevin Drum is confused about the secrecy surrounding "Ida", the name given to the newly-described fossil scientifically known as Darwinius masillae. The fossil represents an unusually well-preserved early primate, and helps illuminate early branches in our family tree.

Drum is surprised that a documentarian associated with the research team working on the fossil said: "There have been lots of reasons for the security and secrecy surrounding this project." Drum wonders:

Hmmm. What reasons? Maybe this is unfair, but something about this reminds me of the fantastic lengths that scholars went to for decades to keep the Dead Sea scrolls under wraps. In that case, it seemed to be motivated by pure professional greed from a group that was determined not to let anyone else contest their interpretations or beat them to a discovery. In this case, it's — what? A desire to wait until a massive publicity campaign was ready?

Drum gets at part of the issue when he adds:

Scientific research isn't all done in the glare of a spotlight, and peer-reviewed research takes a while to finish. Maybe there's nothing wrong with this group's Manhattan Project-esque secrecy. But something about it rubs me the wrong way. Am I off base here?

In short, yes. Part of the reason is procedural. Journals like Nature and Science will not publish research that has already been published elsewhere, and they take an expansive view of what it means to publish. So issuing a press release too soon could count as a publication, and cut you off from publishing in a prestigious outlet. The rules at PLoS are, I believe, looser, but the culture is established, so you keep quiet until you've got your results.

The rules aren't arbitrary. Once research comes out, it's all out in the open and people are free to critique and to reanalyze the specimen. But beforehand, there's a chicken and egg issue. If you just announce that you've got this specimen and invite people to review it, they have to know what it is, and whether they want to look at it. If you say it's a new basal primate, you have to be able to back that up. That means doing research, taking measurements and conducting analyses to show how old the specimen is, and how it relates to other fossils. When a major journal agrees to publish the find, they stake their reputation on its legitimacy almost as much as the authors do. So they need to know that the find has been checked and rechecked, and that the results were obtained through the best scientific practices.

In general, this tends to prevent scientists from hyping their results. That obviously didn't happen with D. masillae. As Brian and Carl point out, the find is probably not quite as important a transitional form as the press releases tend to claim. It is preserved beautifully, with identifiable stomach contents, articulated post-cranial skeleton, and even a furry outline thanks to bacterial action. The paper's phylogenetic analysis is weak, and the public claims about D. masillae's relationship to other early primates are probably overstated. It's a good early adapid, but it probably won't revolutionize primatology.

As PLoS folks like to remind us the real peer review doesn't happen before publication, but after. We're already seeing some of that on the blogs (conversations that formerly would have happened offline, in seminar rooms and at university lunch tables), and in a year or so, we'll have research papers dissecting this first paper. Some of its findings will hold up, others won't. Maybe it really will redraw our understanding of the early branches in the primate family tree. And maybe it won't. That's where the openness of science comes in. The people working on those papers will do their work in private, discussing the results publicly only when the work is done and in press, at which point it, too, will be open for all to see.