I recently sent a request to the editors of several prominent biological, biomedical, and medical journals. I wanted to better understand the current policies regarding access to full manuscript files ? e.g., correspondence with authors, confidential reviewer comments and ratings, and internal editorial exchanges.
My interest had been piqued by our story in the March issue on a unifying metabolic theory.
His files, however, didn?t include internal documents or ?for editors? eyes only? correspondence. These files were off-limits files back then, even to the authors of the manuscripts in question, and even at journals experimenting with ?open? peer review. I wanted to find out whether things had changed, and what the reasoning was for the current policies. Did editors revisit their policies from time to time? And how many requests did they get for this kind of access, anyway? Had a historian, for example, writing about a particular chapter in scientific history, ever requested such a file?
I am pleased to report that, with one exception, everyone responded promptly to the query. The policies were consistent, and in my opinion, flawed. Every journal I contacted has a policy of strict confidentiality; no one provides access to manuscript files under any circumstances. The clarity is admirable, and the practice is so ingrained that none of the journal editors feels the necessity to publish them. No one had received requests for access to files.
Thus, we have the most perfect of secretive regimes: Manuscript files are classified documents, will never be declassified, and the scientists that work under these Draconian prescriptions actively support them, or are at least passively complicit. Under other circumstances we?d be up in arms about our rights. Well, I am up in arms.
The one exception to the ?under no circumstances? dictate, so far as I can tell, occurred last year when Science turned over the complete file for the fraudulent Hwang stem cell paper to an expert panel.
What?s true for the Hwang paper is true for all controversial or celebrated articles: the first cloned mammal, for example, or the sequencing of the human genome, where the reviews, the interactions between the two sets of authors, and the horse-trading between the journals would make for riveting reading. And who remembers the ?memory of water? phenomenon of 1988, when Nature brought in a magician? I?d pay good money to read that file.
So what possible benefit can confidentiality serve? Who is being protected from what? One editor who responded to my e-mail replied that ?review processes everywhere benefit from candor, and abundant experience tells us that without confidentiality evaluators are likely to be less candid.?
If this is true, it can be accommodated by instituting a period of confidentiality for manuscript files. Here?s my proposal: Science journals open their files to reasonable requests after a five-year interval. Just as government files are made public ? in the United Kingdom after a 30-year lag, in the United States after 25, according to change late last year ? this would be a powerful contribution to an open society. It will get to the heart of how research is done and how human relationships govern science. And it will be a goldmine for science history studies, which are not given nearly enough credence.
I?d love a couple of months? sabbatical poking around in the dusty storerooms of the major journals, wouldn?t you?