July 11, 2019 by joshbolick
One of my ongoing research interests involves publisher sharing policies, which stipulate if, how, when, where, and in-what-version authors of scholarly articles are able to share their work. For example, it’s fairly common for the author of an academic journal article to be able to share the accepted manuscript version of their article (not the final published version — gold OA notwithstanding) on their personal website/blog, in a subject repository, or an institutional repository. Sometimes there are embargoes, or delays, on the release of the full text. To date, I’ve focused on Elsevier’s Sharing Policy, partly due to the reaction of open access advocates when it was announced in 2015, and partly because of an inconsistency in the policy which permits immediate sharing of the accepted manuscript on the author’s personal website or blog with a mandated open license (CC-BY-NC-ND) and embargoes of up to 48 months for repository sharing. The inconsistency is that if an author posts their accepted manuscript on their personal website or blog with the open license, then any repository (subject or institutional) has all the permission that they require to ignore the embargo and release the full text of the same version in the repository immediately. Theoretically, the accepted manuscript of every Elsevier article published since 2015 could be shared immediately without embargo, though it’s an admittedly fiddly process. Since my job as a scholarly communication librarian is to support my researchers in sharing their work as openly and quickly as possible, I see this as a tool in my kit to apply when the conditions are appropriate. In 201,7 I presented a poster about Elsevier’s sharing policy loophole, and in 2018 I published an article on the same. To my endless delight, a few folks are calling it “the Bolick Method” or “Bolick Manoeuvre” (I’ve never had a method before, much less a manoeuvre, so I will gloat just a little).
Which is why I was fascinated to recently learn that Cambridge University Press has adopted a broadly similar policy in March 2019. Hat tip to Alessandro Blasetti for alerting me to it on Twitter. In brief, CUP permits immediate sharing in any venue of the pre-print AKA submitted manuscript with the option to apply “any license terms the authors choose,” such as any Creative Commons license. CUP defines the venue choices as personal site/blog, departmental or institutional repository, non-commercial subject repository, and/or commercial repository or social media site e.g ResearchGate, Academia.edu, SSRN, which more or less captures the spectrum of options. For the accepted manuscript, authors are permitted to share “on acceptance” on their personal website or blog, on acceptance for humanities and social science journals, and with a six month embargo from the date of publication for department, institutional repositories, and/or non-commercial subject repositories (no full text on commercial repositories and social media) for science, technical, and medical journals. As with the pre-print, authors have the option of applying a CC license, though here they’ve restricted the choice to CC-BY-NC-ND. Similar to the Elsevier policy, the open license, if applied, permits the immediate sharing of the accepted manuscript, nullifying the six month embargo required for STM titles. There’s a handy table, “What can be archived, where and when,” summarizing this on the CUP green open access policy site. Note that the terms of this policy apply retrospectively: “if an author has any Green OA conditions in their journal publishing agreement that are more restrictive than our new standard policy, then our new standard policy supersedes the relevant items in their publishing agreement.”
So while CUP has provided a way for authors to undermine the embargo, a la Elsevier, there are some notable differences:
- More freedom re: license; authors are permitted but not required to apply an open licence. For preprints, they may select any license they wish, including licenses that permit modification. For accepted manuscripts, they may only apply CC-BY-NC-ND, which is the most restrictive CC license, but any CC license is more open than all rights reserved.
- No embargoes for humanities and social science journals; pre-print and accepted manuscript publicly shareable from acceptance everywhere except commercial sites.
- Short embargoes for science, technical, and medical journals; 6 months from date of publication.
- Retroactive; increased open opportunity for articles previously published under more restrictive terms.
To be clear, I’m not critiquing CUP’s new sharing policy the way I clearly was Elsevier’s. For the most part, there’s a lot here to feel good about. CUP should be commended for extending greater choice to authors in license terms, the elimination of embargoes for some titles and relatively short embargoes for others. If other publishers would follow suit, we would have a better scholarly world, at least in terms of author’s rights, reuse rights, and access. That said, what are the deeper implications of these particular points of CUP’s sharing policy?
First, the option to apply any (or no) CC license to the pre-print is a positive step towards greater author choice and freedom. I wonder: could not a motivated party (assuming a license that allows modification) leverage the CC license to develop the pre-print to match the content (including general format and pagination; excluding any trademarks) of the final published version of record? The application of an open license may be particularly important for articles with visual elements (images, charts, tables, graphs, etc.), assuming those elements are subject to the license, for future reusability by both the author as well as any reader/researcher who may benefit from reuse of those elements. More flexibility for authors for the accepted manuscript would be desirable but given the ability to manipulate the pre-print maybe it’s not important? In any case, as long as any permitted CC license is applied to either the pre-print or the accepted manuscript, they are immediately shareable via repository indirectly vis-à-vis the CC license rather than directly through the sharing policy.
However (getting to the second point), the motivation to leverage the CC license to subvert an embargo is entirely eliminated for humanities and social science journals, given zero required embargoes, and severely blunted for STM titles where the embargo is a mere six months. IMO any delay on sharing is an unnecessary and in at least some cases arguably or actually harmful (a number of STM issues come to mind), and the time from acceptance to publication can be quite long, which is effectively extending the embargo (X months from acceptance to publication + 6 months) when compared with titles that permit sharing upon acceptance. So it could be improved (no embargoes, period), but it’s better than industry standard (12+ months). If sustainable, this policy further undermines the claim that embargoes are necessary at all, for which there is little evidence, and which runs counter to a growing trend (see “Open access: ‘no evidence’ that zero embargo periods harm publishers” by Rachael Pells, Times Higher Ed, April 23, 2019). As I’ve argued elsewhere, the haphazard availability of pre-publication manuscripts on the open web isn’t a factor in library subscription decision-making, and there are MUCH bigger threats that should take higher precedence anyway.
Considering the implications of no embargoes in some disciplines (in this case, HSS), and embargoes in others (in this case, STM), what does that mean about the differing implied value of knowledge in those disciplines? Obviously, I reject the idea that free = lesser value or quality, but there’s something going on with free X and restricted Y alongside each other. If in the school cafeteria I freely give my grapes away but hold back the brownies, does it not say something about my evaluation of the differing desirability of both? That’s an admittedly simplistic analogy and economics is vearing way out of my expertise, so I’ll stop there. It’s interesting that CUP’s general take here, no embargoes for HSS, 6 months for STM, inverts Taylor & Francis’ model: generally 18 months for HSS titles, and 12 months for STM titles (per SHERPA/RoMEO).
So…takeaways: kudos to CUP for a sharing policy that is generally progressive; IMO these are clearly positive steps towards greater openness and rights retention for authors. In my back and forth with Elsevier regarding the loophole, I suggested that maintaining the open licenses but eliminating embargoes would be a significant step forward and demonstration of leadership, and here CUP has almost fully done so. That should be acknowledged. Authors and repository managers: take advantage of these rights to ensure broader access, visibility, and impact of scholarship published in CUP journals.
Let me know if I’m missing something crucial! If anyone at CUP would like to comment on these aspects of their policy and why they chose to implement them, I would be happy to incorporate it into this post!
h/t to Maria Bonn for her very helpful suggestions and questions on a draft of this…essay?
-Josh / @joshbolick