When it was announced that the faculty at Princeton University had unanimously adopted an open access policy for scholarly articles they authored, it was great news for the open access community, but it was also the cause of some overheated rhetoric. Since the operative language of the Princeton policy differs very little from that that was adopted at Duke back in March, 2010, this is a good opportunity to reflect on what has, and has not, been done.
In all such policies the university is given a license in the works that is prior to any copyright transfer to a publisher. Technically, therefore, the rights that are transferred are subject to that license; hence the language of “banning” the wholesale transfer of copyright, which has received a lot of attention. I wanted to point out, however, that this rhetoric about a “ban” did not come from Princeton itself, but from a single blogger, to whose post all the stories that use that language point. That blogger has now changed the post, including a quote from a Princeton official saying that the faculty is not being “banned” from anything. Even the URL has changed; the corrected version of the post is here.
The differences amongst universities regarding these policies come in implementation. Some universities may elect to act in a way that is contrary to the terms of the publication agreements the authors enter into (by posting articles or versions of articles where the publication agreement purports not to permit the specific posting). Doing so would seem to be legally permissible under the claim of a prior license, but it could also put the faculty members in a difficult position unless they are very careful about what they sign (as they should be but seldom are). An alternative is for the university to exercise the license in a more nuanced way, taking into account the various publisher policies as much as possible. That, of course, makes open access repositories much more labor-intensive and difficult, especially as publishers change their policies to try a thwart these expressions of authorial rights. How Princeton will actually implement its policy is still an open question, since they do not yet have a repository of their own.
Earlier today I received an inquiry about the Princeton policy from a colleague at another university. To what degree, he asked, is this similar to the university simply claiming that scholarly articles are work made for hire? My answer, of course, was that these policies are the very opposite of an institutional claim of work for hire. If that were done, in fact, no such license would be necessary. But these policies are founded on faculty ownership and express the desire of a faculty, as copyright owners, to manage their rights in a more socially and personally beneficial way. It is important to note that the open access policies now in place at a couple of dozen U.S. institutions have all been adopted by the faculties themselves; they decided to grant a non-exclusive license to the university, which, again, they could not do except as copyright owners.
Probably the most important fact about these policies, indeed, is that they represent an assertion of authorial control. We so often hear publishers and others in the content industry talk about protecting copyright, by which they usually mean the rights they hold by assignment from a creator, that it is salutary to remind academics that they own copyright in their scholarship from the moment their original expression is fixed in tangible form. Transferring those rights to a publisher is one option they have, and it has become a tradition. But it is only one option, and the tradition is beginning to be questioned, as this recent article from Times Higher Education and this one from Inside Higher Ed forcibly demonstrate.
Open access policies are not, at their root, either “land grabs” by institutions or acts of defiance aimed at publishers. They are simply a recognition of the fact that authors are the initial owners of copyright, and they express a desire by those owners to manage their rights intentionally and in a way that most clearly benefits the goals of scholarship.
Like its Harvard model, Princeton Open Access Policy needs to add immediate-deposit requirement, with no waiver option
1. First, congratulations to Princeton University (my graduate alma mater!) for adopting an open access mandate: a copyright-reservation policy, adopted by unanimous faculty vote.
2. Princeton is following in the footsteps of Harvard in adopting the copyright-reservation policy pioneered by Stuart Shieber and Peter Suber.
4. I hope that Princeton will now also follow in the footsteps of Harvard by adding an immediate-deposit requirement with no waiver option to its copyright-reservation mandate, as Harvard has done.
5. The Princeton copyright-reservation policy, like the Harvard copyright-reservation policy, can be waived if the author wishes: This is to allow authors to retain the freedom to choose where to publish, even if the journal does not agree to the copyright-reservation.
6. Adding an immediate-deposit clause, with no opt-out waiver option, retains all the properties and benefits of the copyright-reservation policy while ensuring that all articles are nevertheless deposited in the institutional repository upon publication, with no exceptions: Access to the deposited article can be embargoed, but deposit itself cannot; access is a copyright matter, deposit is not.
7. Depositing all articles upon publication, without exception, is crucial to reaching 100% open access with certainty, and as soon as possible; hence it is the right example to set for the many other universities worldwide that are now contemplating emulating Harvard and Princeton by adopting open access policies of their own; copyright reservation alone, with opt-out, is not.
8. The reason it is imperative that the deposit clause must be immediate and without a waiver option is that, without that, both when and whether articles are deposited at all is indeterminate: With the added deposit requirement the policy is a mandate; without it, it is just a gentleman/scholar’s agreement.
[Footnote: Princeton’s open access policy is also unusual in having been adopted before Princeton has created an open access repository for its authors to deposit in: It might be a good idea to create the repository as soon as possible so Princeton authors can get into the habit of practising what they pledge from the outset…]
Stevan Harnad
EnablingOpenScholarship
Kevin, I think that we disagree about the Princeton announcement. My reading is that it is a substantial advance over the policies adopted by Harvard and Duke, and it presents a different licensing challenge than you suggest.
The Princeton policy is important in that it seeks to provide access to research as published, and not just to the “Accepted Manuscript” (as defined by the Journal Article Versions standard). It allows authors to deposit a version prior to copy-editing and other tweaking by the publisher, but only as a matter of convenience. The preferred version will be the version as published.
In order to accomplish this, a license granted to the university prior to a copyright transfer to a publisher will not be enough. The author cannot license to the university the work that the publisher will contribute to the improvement of the article. Princeton faculty instead must promise the university that they will secure from the publisher permission to post in an open access forum the article as published. And in their negotiations with the publishers, the faculty must secure those rights, perhaps by using the SPARC Authors Addendum which guarantees that faculty can post the published work.
The challenge of course is what will happen if the faculty renege on their promise to the university. The university might post to an open access repository the published version of a faculty paper on the assumption that the faculty member has secured the necessary rights. If the faculty member has not done so, would the university then be guilty of copyright infringement?
The Princeton initiative nevertheless represents an important step forward in the development of open access “mandates.” Initiatives such as Harvard’s merely require that faculty members inform publishers that they have already partially-licensed the rights to manuscript versions of articles. The Princeton initiative requires faculty to secure from publishers the authority to distribute copies of the articles as published. It may be hard work, but as long as we believe that publishers are improving scholarly writing, it is the only option that is worth substantial effort.
Thanks, Kevin, for an interesting commentary on the Princeton policy.
Peter – I am not sure about your comment that this policy requires the published version to be made OA? I have read through the policy and cannot see where this is claimed. The closest I can see is:
“Q. What kinds of publications are covered by the policy? A. Refereed journal and conference articles actually published; not unpublished drafts, not books, fiction, poetry, music, film, lecture notes, case studies, etc.”
The reference here to “actually published” surely stands in opposition to unpublished drafts, fiction, lecture notes etc. In other words, articles that have been “actually published” rather than articles “as actually published”. Or am I missing something?
Telling authors to retain copyright over their own work seems fairly straightforward. Asking authors to obtain, rather than retain, copyright in the published version will lead to the complexities of intention and implementation that you mention and would, I suspect, deter authors from having anything to do with the policy.
Bill Hubbard
Centre for Research Communications
Bill, I was basing my comments on the statement that faculty license to Princeton the right “to exercise any and all copyrights in his or her scholarly articles published in any medium…” (emphasis mine). Earlier, the background to the policy states that its purpose is “to make the faculty’s scholarly articles, published in journals and conference proceedings, available…” In other words, they are not talking about accepted manuscripts. Most telling of all is the final entry in the FAQ, which I repeat here in full:
<
I find two things about this question notable:
1. It recognizes that most faculty want the final, corrected, published version of their work to be available, and not the “accepted manuscript” version favored by so-called “green” open-access initiatives.
2. Depositing the accepted manuscript is not the preferred method, but an acceptable alternative.
Kevin and Bill, let me know if I am misreading this policy.
Peter – you base your claim on: “to exercise any and all copyrights in his or her scholarly articles published in any medium…” and – ok – I see your point.
I am not sure I agree though: I read this with the sense of: “to exercise any and all copyrights in his or her scholarly articles [that he or she have written and finalised as articles, that then go on to be] published [published, as opposed to being grey or draft articles that never make it to publication] in any medium…” Kind of clumsy to state it that way, but do you see the interpretation?
The same sense can also hold true for the Q&A section, which to me reads like they are trying to make things easy for the author by just pointing out that some publishers allow author’s final version distribution as part of their existing policies.
Your further points:
>1. It recognizes that most faculty want the final,
>corrected, published version of their work to be
>available, and not the “accepted manuscript” version
>favored by so-called “green” open-access initiatives.
I think its not so much that green OA *favours* author’s final versions, but that this is often the only thing that publishers allow authors to use. Some publishers allow and some publishers insist on the use of the final publisher pdf: but many don’t allow it. Of course, in terms of the content, then it should be the same: if there is anything significantly different, then the sub-editor should go down as a joint author ! 🙂
> 2. Depositing the accepted manuscript is not the
> preferred method, but an acceptable alternative.
I agree that I take from my reading that the author’s final version is acceptable. Many academics do prefer the publisher’s final version, but whether this is a neccesary preference, or one that is from inertia of habit remains to be seen. And of course, the many researchers without current access to the published version are very happy indeed with the author’s final version through OA.
As far as which interpretation is intended – I will be interested to see how this develops in use. I suspect that if the author’s final version is acceptable, then this will become the standard method of compliance through ease of practice, whatever the original intention.