The mandatory NIH public-access policy goes live today. The sky does not seem to be falling (I just checked), but you wouldn’t know that from some of the reactions I’ve heard. Since this week has been declared OA Week, I thought I’d kick it off by retailing some of the things I’ve heard that worry me, with all appropriate serial numbers filed off.
I haven’t heard any worry over the fate of publishers. If I were a toll-access publisher, I would be worried by this. As I’m not, I’m not.
Without undue tooting of my own horn, let me say that my concerns about recalcitrant publishers have proven to occupy a lot of faculty brainspace. I don’t live in the medical-research realm, so I don’t know how much of this worry is futile handwringing and how much identifies a genuine problem. I only know that deans are worried particularly about protecting their junior faculty, who already find publishing an uphill climb. The sooner we all address this, the easier we will all find the compliance process.
I have heard a lot of worry over the versioning problem, from faculty spanning quite a few disciplines (with the understanding that “NIH grantee” implies a fairly narrow range to begin with). “What happens when copyediting catches real errors, or changes the thrust of an argument?” runs the basic version of this question. “The version in PMC will simply be wrong.”
This is not a silly or uninformed objection. I used to work for a publishing-services bureau; I know full well what “final” peer-reviewed manuscripts look like before and after copyediting. Let me tell you, a good copyeditor is worth his or her weight in diamonds, and it warms my text-artisanly heart that faculty realize this as well. Final manuscripts can be disasters; research faculty can be terrible writers and terrible writing teachers.
I saw a suggestion that copyediting happen before the journal-submission process. I’m sorry, but no, that won’t work. Part of the copyediting process involves bashing the manuscript (citations especially) into journal style, which obviously can’t be done until the journal has been chosen. Another part is marking it for typesetting, which is intimately connected with each journal’s particular typesetting practices. You can’t just pick copyediting up and dump it in front of manuscript submission; you’d only add a step, slowing the whole publication process down and increasing its cost.
At present, the only workaround for this (as I understand matters) is working with a publisher cooperative enough to replace PMC’s manuscript version with the published version. These publishers exist, but they are not exactly numerous. For PR purposes if for no other reason (and “accuracy” is a plenty good enough reason all by itself), I think it would be wise for PMC to work out a way for PIs and other authors to fix errors in their manuscripts. I have heard the versioning problem called “a flaw in the policy” and “suicidal” by people in very high places.
Another difficulty has to do with the principal investigator’s responsibility under the policy, given that the PI is likely not an author (much less the first or corresponding author) on every single article coming from a given NIH grant. This is a tough one to resolve, given that the buck has to stop somewhere, but I would suggest at the least that first/corresponding authors as well as PIs be able to approve manuscripts and offer corrections.
Reading the NIH’s comment stream, I see that the too-much-work backlash has begun. In my cynical way, I tend to ignore this particular objection (trusting in Swan’s research on mandates to back me), but if we open-access advocates want to be smart about this, we will sort out how to help libraries offer third-party PMC submission services. (Disclaimer: MPOW does this. It was not my decision, though I support it wholeheartedly.) PMC could help us all by providing a deposit API (preferably based on SWORD) that those of us with institutional repositories could program against. Not only will that allow people like me to get in on the repositing action, it will help institutions monitor compliance and provide useful services (such as local PMCID/NIHMSID databases) to faculty. As mandates become more numerous, local services become even more important, as they allow faculty to become accustomed to one deposit interface, not a dozen. Please, PMC, set the example here!
Confusion about the policy is rampant, though MPOW is doing its level best to calm the waters. I do worry about institutions that are letting this slide—which is most of them! I suggest that the NIH target marketing efforts at libraries as well as researchers. MPOW and other institutions like it prove that we can help. The problem is that not all of us know we should. If the NIH isn’t working with SPARC to clue in the larger research-library community, it should be.
Happy Open Access Week!
