We are trained these days to dread “unintended consequences” whenever we make decisions; it is a fear that sometimes leads to paralysis. But not all unintended consequences are negative, and I want to take a moment to celebrate some unexpected things that have resulted from the adoption of an open access policy by the Duke University faculty back in March.
The two biggest consequences so far have been a flurry of activity in the Duke Libraries and some welcome attention from outside the University.
In the latter category, this interview with Paolo Mangiafico, which was posted recently on opensource.com, is a superb summary of the rationale behind the policy and the steps we are taking to implement it. Paolo, in his rather unique position at Duke as Director of Digital Information Strategy, provided the leadership that was necessary to organize, draft and advocate for the open access policy. Paolo himself is a unique combination of high-level IT skills with a deep understanding of policy options and consequences, and his description of our OA policy is as articulate as any I have seen.
In the interview, Paolo talks about the decisions that must now be made, both as a matter of technological infrastructure and in terms of re-imagining library services. Thus his interview nicely encapsulates the reasons behind that flurry of activity I spoke of above.
Another particularly exciting consequence for me personally has been an invitation to speak at the the 8th Berlin Conference on Open Access on the legal issues involved in open access. The invitation letter clearly indicates that Duke’s new policy is one of the reasons for this opportunity, and I am honored to be invited to this influential gathering, which drafted the Berlin Declaration on Open Access back in 2003. In 2010, the Berlin Conference will be held in Beijing, China, which deepens my excitement, both because I have never been to China and because of the opportunities the conference offers to learn about the progress of the open access movement in Asia (although the activities of SPARC Japan are already well-known). I hope I shall see many friends from North America and Europe in Beijing, and I look forward to the opportunity to meet many new colleagues and friends.
Sometimes unintended consequences offer really delightful surprises.
I had been hoping that you were going to write more about Duke’s Open Access policy, especially since you were so eloquent in your arguments on why all universities don’t need an institutional repository. As a personal matter, I think retaining rights and providing open access is the only way to go. But I have been having a harder and harder time understanding why librarians as a group would support it. Sure, as a matter of principle librarians are in favor of the broadest possible dissemination of information. But open access mandates such as Duke’s will only provide access to an inferior product (an author’s manuscript, an item that university archivists routinely discard as being of little research value), and at a substantial cost in terms of software development and library support (as Paolo Mangiafico makes clear in his interview). Furthermore, there are no cost savings associated with open access. Again, as Mangiafico states, this is all an addition to everything that Duke is doing and buying, not a replacement. While he calls for a new business model for scholarly communications, it is not clear that an open access mandate actually contributes to the development of such a model.
So on a very practical level, what are the advantages of an open access mandate for a library (other than a personal invitation to speak in China – congratulations!)? Has the administration at Duke, for example, provided new positions and new funding in order to support this new supplemental distribution medium? Or do librarians now feel that it is more important to spend money to provide free access to everyone to unedited manuscripts than it is to provide limited access to the published record?
I think there are some unfortunate and erroneous assumptions in this comment, two of which I would like to address.
First, I think the distinction between the author’s final manuscript, which has been through peer review, and the published version is not as stark as Peter suggests. Whatever editing occurs as that manuscript is prepared for publications should not, as I have argued before, involve substantive changes without input from the author. If such changes are made, it is the published record, not the author’s final manuscript, that should be suspect. In any case, in many disciplines in which research is time sensitive, researchers are already relying on pre-publication manuscripts, either using disciplinary repositories or informal exchange of manuscripts. I am quite surprised to learn that archives would discard such material, because I am convinced that if libraries remain implacably wedded to only the published record, they risk increasing irrelevance.
Second, support for an open access initiative by libraries is not entirely about what we as a library gain. Our faculty and our institutions will reap benefits from open access, as Paolo articulates in his interview. Our fundamental calling is to serve the mission of the institution, and support a repository is an important part of that task in the digital environment. We will be reassessing some staff roles in order to support IR services, and much of the infrastructure work is already going on. We do not plan, I would emphasize, to reduce collection budgets. But we cannot ignore the benefits to our researchers and to the institution that an OA repository offers. This is especially true at a university where all three of our strategic themes — internationalization, interdisciplinarity and knowledge in the service of society — will be advanced by this initiative.
I should add, on a personal note, that the piece Peter cites in which I argue that not all institutions need have an institutional repository was originally written as part of debate for the chapter of ASIS&T at UNC SILS, and I was assigned the negative position. I am glad Peter thought it was articulate, since I was making an argument I was asked to make and it is a challenge to do that well. In any case, that piece does not suggest that IRs have no value, but that some institutions, usually smaller ones, may be best served by using disciplinary repositories or consortial arrangements to reap the benefits of OA.
Thanks for your comment, Peter. Just before posting this comment, which I wrote offline, I noticed that Kevin had just responded and addressed some of the points I planned to make, but I’ll go ahead and add mine to provide my perspective, not as a librarian but as someone who has worked closely with librarians for many years and is now looking at these issues from the perspective of a provost’s office.
The simple answer is that the open access policy that Duke has just adopted was not intended to provide tactical advantage to the library, but to serve the strategic mission of the university. The library happens to be one of the organizations on campus best equipped to meet the university’s goals of organizing, sharing, and preserving the fruits of the research being done here (the university press being another), and librarians happen to be among those who know the most about these issues and are most passionate about promoting them. What’s in it for them? They want to do the right thing, and are willing to take a long term view and to make small investments in the short term with the intention of contributing to a broader movement whose whole may turn out to be greater than the sum of its parts.
With respect to the questions about costs, “inferior products,” adding on to the existing workload rather than replacing something, and potential impact on changing publishing models, one could ask the same questions about projects to digitize manuscripts or other unpublished materials from library collections (or for that matter, of any website development project for any organization on campus). Yet many libraries have built digital collections programs and dedicated significant resources to them. What are the practical advantages to libraries of doing that? (I ask this rhetorically – I’m a big supporter of library digital collections programs, having been involved in the earliest digitization projects at Duke Libraries in the mid 1990s as director of the Digital Scriptorium). The costs of providing services around open access to faculty research are relatively small compared to what we’ve invested in digital collections, and we’re planning to build on and leverage the investments made in the digital collections program to support these other goals too, so they won’t all be new costs.
And do we really want libraries to limit themselves to “provid[ing] limited access to the published record”? I hope that libraries will continue not only to explore how their roles can grow and change in response to transformations in the scholarly communication ecosystem, but also to help shape the transformations by doing things that promote the values and strategic mission of the university in the areas where they have the greatest expertise, even if it’s sometimes difficult to see how it might provide short term tactical advantages.
Thanks Kevin and Paolo for the responses. I think that we are in agreement that it is part of academia’s mission to make the product of its researchers readily available to the community of scholars (and by extension, the general public). I am a big supporter, therefore, of initiatives that encourage faculty to retain copyrights and make material available on their own terms. I also believe that those who can afford it should have IRs, and I don’t have too much of a problem with a library running one, since it is often relatively cheap (though Cornell’s experience trying to maintain arXiv is making me wonder about that a little).
I do believe, though, that librarians have made a tactical mistake by accepting the author’s Accepted Manuscript as the unit of choice in initiative like PUBMED and FRPPA. In some fields, the differences between the author’s Accepted Manuscript and the final version are trivial. Some publishers even demand camera-ready copy from authors. But in other fields, the contributions of reviewers and editors are substantial. I know with my own writing, the acceptance of the paper is usually only the first step in a long process of substantial changes as I first respond to the reviewers’ comments and then react to suggestions from the editor and copy editor. The NISO Journal Articles Version standard seems to assume that all of the changes are made prior to acceptance, but in my experience as an author and reviewer, the “accepted but revisions required” status is much more common.
But even if we assume that the differences between the Accepted Manuscript and the Version of Record are minor, I am still troubled by librarians ceding their traditional role of managing and preserving the Version of Record to publishers. It is this item that we want to have in our repositories, not all of the drafts that led to the Version of Record. In a relatively small number of cases (think Watson’s and Crick’s article on DNA), historians may want to follow the editorial process that led to the Version of Record, but few academic authors are modern Shakespeares or Joyces, and the versions of the paper that lead up to the Version of Record are not worth saving.
This distinction is also why I don’t think the digitization example holds. The primary sources that are found in libraries and archives are rich resources for further academic study and research. Authors’ Accepted Manuscripts do not have the same research potential, and it is not worth spending library resources to collect and preserve them (unless it can be done very, very cheaply).
I am also not convinced by the argument that just because it is part of the University’s mission, the Library should assume responsibility. I think that initiatives like MIT’s OpenCourseware are just as important (if not more) to the University’s mission, and librarians would be just as capable at “organizing, sharing, and preserving” the fruits of the teaching being done at the University, but no one is suggesting that the OpenCourseware initiatives should be coming from the library’s budget.
So my bottom line: I hope that either a) Duke can figure out how to run a campus-wide program on the margins, at little cost, or b) the University recognizes that the tasks that it is assigning to the Library are new ones that are not necessarily more important than what the Library is already doing, and therefore need to be separately funded.
Hi Peter. Thank you for the clarification of your comments. I still think your distinction between an accepted manuscript and the version of record is too stark; there is something in between. Your comment assumes that the our repository will contain accepted manuscripts that are deposited prior to the peer-review process (so-called “pre-prints”). But in fact the policy assumes that the default will be the final author’s manuscript, which would include all changes made based on the peer-review process. We would love to include as many final published versions as we can and will work to make that happen whenever possible. But I continue to believe that the differences between the final author’s version (as opposed to the accepted manuscript) and the final published version will not be large or substantive, for reasons I have already stated. In any case, it will be our faculty authors who ultimately will decide when and if a manuscript represents their work sufficiently for inclusion, and they are in the best position to make that judgment.