Dr. Alondra Nelson graciously responded to our queries about the new US OSTP memo concerning public access to the results of federally funded research, and on Tuesday we published her responses in full. After rounding up some reactions to the memo and with the benefit of Dr. Nelson’s responses, we’d like to offer some thoughts of our own. We’ve focused here on the issues that we identified as critical – in fact, so important that we made them the basis of our questions to OSTP.
The shift from mandatory (the language of the Holdren memo) to recommended public access (the language of the Nelson memo).
(Rick) I was disappointed that Dr. Nelson initially declined to answer our question on this topic directly. However, by simply referring back to the language of the memo (repeating the word “should”) and then pivoting to a discussion of how the memo was developed, she seemed to confirm what I originally believed: that the change in language is intentional, and that this new memo represents only a recommendation where the old one represented a mandate. To be clear: if policies mean what they actually say, then the new OSTP memo represents a set of recommendations rather than requirements. This, it seems to me, is a very big deal. And now my questions are:
- Why the shift from mandate to recommendation?
- Will that shift be acknowledged in the scholcomm community?
- Will funding agencies actually treat these new terms as optional?
(Karin) Who doesn’t want openness and accessibility?! But as we’ve long discussed, it’s how you get there and at what cost for whom that matters. I wondered whether this language is meant to provide more flexibility or is simply making assumptions about the direction of travel for open access. Certainly some responses (like from SPARC) seem to read the memo as mandating rather than recommending.
(Rick) Agreed, but I seriously wonder to what degree such public interpretations on the part of advocacy groups represent a good-faith interpretation, and to what degree they represent a desire to shape the public narrative – like football players frantically signaling the preferred outcome of a contested call – and maybe even to distract from the reality of a policy that has suddenly become much less directive than it used to be. I certainly can’t find anything in the language of the document itself to support the belief that it represents an actual mandate, and I find plenty in it to support the belief that it represents a nonbinding recommendation. Dr. Nelson was given the clearest possible opportunity to correct this impression, and chose not to. Again, this is big news, though it’s not clear how much attention it will get.
(Karin) I hope it is offering maximum flexibility to agencies that serve widely different research communities. My principal concern here is with the National Endowment for the Humanities. The NEH has a tiny fraction of the budget of the National Science Foundation ($275m vs. $10.3b – and I can’t believe we think the humanities are only that fraction as important as science to human knowledge and wellbeing). The way that the NEH and NSF, just to choose two agencies of many, fund research is dramatically different. The “outputs” are dramatically different. I hope that this difference is being attended to rather than flattened out as so much monolithic open access policy has been in other contexts.
(Rick) We should note here that while in the process of composing this post, we received some follow-up communication from Dr. Nelson and her Office on the evening of Tuesday, 11 October. This led to a brief exchange in which the Office confirmed that the guidance document does, in fact, represent a non-binding set of recommendations, not a mandatory directive. As I said above: this is a very big deal.
OSTP’s view on whether open science creates any risks as well as benefits
(Rick) Dr. Nelson simply declined to answer this question at all. Instead she responded by delineating the benefits of open science. Again, this is disappointing, and it’s the kind of answer one would expect more from an advocate than from an analyst. In the wake of the COVID crisis in particular, however, such a position seems completely untenable. While there were undoubted benefits to both speeding up the generation of COVID-related science and opening up access to it during the crisis, there were also many examples of greater speed and openness creating unintended and potentially harmful consequences. In a highly complex ecosystem like scholarly communication, disruptive change (such as a sudden shift to highly open research and publishing practices) will always bring a mix of positive/intended and negative/unintended consequences, and it seems to me that to pretend otherwise is both unrealistic and – especially where public health is concerned – dangerous.
(Karin) I agree that to see only benefit is to actually take on even more risk. That is, if you’re not allowing for risk you’re not working to mitigate it either. Some of the risks I see concern the continued reliance on volunteer labor – not to review scholarship, but to edit and publish it – in an environment for the humanities that is increasingly bleak. My own view is that publishing, like health care, should almost always be non-profit. But there are real costs to employing highly skilled people to do the work that takes research into publishable — meaning legible and digestible and thus actually useful – form. I am a huge admirer of Dr. Nelson. Her dedication to equity in science and her leadership both as a scholar and in her previous and current roles are admirable. But I think the potential barriers that open access policies can create among researchers and ultimately for the public we all want to reach are very real and concerning.
(Rick) This continues to be one of the most difficult and contentious topics in scholarly communication: if we’re not going to support the legitimate costs of publication by charging for access to content, then what should we charge for and who should pay – and by what mechanism(s)? Also, which costs of publishing are “legitimate”? There’s a pretty wide range of viewpoints on that question alone.
(Karin) So much of this, I think, stems from different perspectives and positions. People understand the field or discipline or professional position they’re in, but not necessarily any others – and that can lead to serious misunderstanding of how research is funded and what publishers do, among other things. Those “legitimate costs” debates, for example, might refer to dramatically different kinds of and scale of costs. I reviewed the list of OSTP’s external stakeholder consultations that Dr. Nelson referenced in her response to us. I saw commercial publishers, science society publishers, and some infrastructure groups, among others, but I didn’t see university presses – just to take one example of nonprofit, highly skilled human labor costs of publication.
What Dr. Nelson and her team see as the most important and meaningful benefits of open science
(Karin) I assumed they would point to the rapid development of the COVID vaccine as a benefit of open science, but I think there is a lot to Dr. Nelson’s answer that I’d like to hear more about. Three assertions in particular stand out: 1. Openness advances “all areas of knowledge.” 2. When research is widely available to other researchers and the public, it can provide policymakers with tools to make critical decisions. 3. Open research meets the requirement for government accountability and transparency. I want to know how all three of these apply to humanities scholarship. It is not enough to refer to science and be done with it. I don’t think “all areas of knowledge” is in this memo by mistake.
(Rick) That’s a good point, and one I hadn’t considered. Embarrassingly, it hadn’t fully registered with me that the White House Office of Science and Technology Policy is issuing guidance that applies to funding for disciplines well outside the realms of, you know, science and technology.
Potential impact of the policy changes on non-profit scholarly publishers
(Karin) If these are recommendations only, then that assumes agencies are free to adapt them to their research communities in ways that best meet the challenge of public access. As we’ve noted time and again – and as Robert wrote about mathematics last week – if your discipline is not primarily serviced by large volume commercial publishers, it likely means a couple of things: society publishing, other nonprofit organizations, and generally low cost research and publishing. We should be encouraging rather than creating more barriers to exactly this model.
(Rick) The good news here is that Dr. Nelson and her Office have now made it clear that this guidance is, in fact, a nonbinding set of recommendations, and that agencies are in fact able to adapt them as needed. What I find interesting – odd – is that although the Holdren Memo was written as a mandate, its terms were also explicit about allowing a certain amount of flexibility in implementation, and about the need for agencies to work in consultation with other stakeholders, including publishers. The Nelson Memo is written as a recommendation only, but at the same time has stripped out all the language about discipline-specific adaptations and consultation with publishers and disciplinary organizations.
(Karin) I see lots of challenges for the humanities but want to point to what I think is by far the most important for this policy or any humanities funder open access policy or guidance: what funding threshold triggers the policy? The National Endowment for the Humanities (NEH)*, for example, funds summer research for faculty, or year-long support for lots of scholars including those doing commercially published work. But neither the summer funding or the year of support would constitute the majority of funding needed to write a book in my field (history). In the case of a prolific scholar, for example, who published a book last year, a year of ACLS funding, a year of NEH funding, university funding, and multiple years of research trips including small grants from research libraries all contributed to the final work. Would the NEH funding –federal funding – trigger the public access requirement? The memo, because it is science-driven, refers to journal articles. But that is not the apex publication in many fields. But even so, you would have a hard time sorting out which portion of a history journal article’s research was supported by the NEH as opposed to other funding sources or unfunded time. Would the scholars funded under the NEH public scholars program specifically designed to support public facing work be required to publish in open access rather than in the trade presses that give their work more potential for a general audience? I think not, and rather that the NEH should be able to make some requirement for a public-facing and freely available component to the scholars work. I understand that kind of idea cuts against the argument that it is the scholarly version that should be available to all, but I think it meets the basic principle of public access – and for the public good.
Whether the posting of preprints would constitute sufficient compliance with the new policy
(Rick) The response to this question in our interview reiterated one element of the new guidance that I had initially overlooked: that for deposit to be compliant with the policy, it must be in an “agency-designated (repository).” This would indicate that only a subset of institutional repositories are considered an acceptable locus for reports on funded research, and that those repositories will be selected by the funding agencies, not by the authors. This has potentially serious implications for the ongoing viability of institutional repositories of the kind hosted by college and university libraries. Of course, since OSTP has confirmed that the whole guidance document constitutes a set of recommendations rather than requirements, this point may seem to be mostly moot. But that only applies at the level of the OSTP guidance itself; in response to the guidance, funding agencies are going to set rules that absolutely are binding on those whose work they fund. So if Agency A says “future funding from us will be contingent on factors that include the placement of your work in one of the repositories on this list,” that will certainly be a binding condition.
(Karin) Last week we noted that Brett Bobley, CIO for the National Endowment for the Humanities and the Director of the Office of Digital Humanities, responded to a query on Twitter about APCs and repositories with a reference to (and a link to) Penn State Libraries guide to green OA. He noted that NEH would have to “ work closely with OSTP and other agencies to figure out an actual policy. But it sure sounds like green OA.” The Penn State guide refers to the university’s repository: “ScholarSphere is Penn State’s institutional repository — a repository service for research and scholarly work produced by the university community. Through ScholarSphere, Penn State researchers — faculty, students, and staff alike — are able to share their work on a worldwide scale and be assured of its long-term preservation and thus ongoing access.” Will such library repositories be “agency-designated?” If they are, that comes with its own challenges. The costs to universities of these local repositories, and the challenges they pose for discovery, have been widely discussed. But if they are not, would an agency be inclined to instead create its own – or designate a commercial repository? I agree this may be moot. But I also think that like with many unintended consequences of open access policies this really bears a long, close look.
Karin and Rick welcome other thoughts, observations, and analyses – send in your comments!
* (Karin) Full disclosure (and public information!) I am the PI on an NEH grant for fellowships from the library I direct, and I have been the PI on the same type of NEH grant in my previous position.