Stevan Harnad posted a glowing encomium for Poynder’s article to the JISC-REPOSITORIES mailing list. This is no surprise. What surprised and gratified me were two cogent, politely critical responses, which Harnad has as yet not answered. Perhaps there will be more criticism; perhaps Harnad will respond. We shall see.
Let me be clear: what Poynder said about librarian open-access efforts was offensive and shortsighted, and he could do worse than apologize for it on his weblog. Harnad’s enthusiastic response to Poynder tars him with the same brush, and he could stand to apologize also.
But just for fun, let us play out a Harnadian scenario, in which libraries use IRs for their own projects and OA happens somewhere else entirely. After all, Harnad is entitled to throw spaghetti at the wall too. If his spaghetti sticks to the wall, then libraries win. If I end up out of a job because Harnad’s spaghetti sticks—well, even I consider that acceptable collateral damage. Information to the people!
Harnad’s own vision (based on his email, which is repeated at this post to his blog) appears to be small departmental faculty-spurred and faculty-owned fiefdoms, which would then be aggregated at the university level for harvesting and dissemination purposes. This would indeed carry some advantages: faculty evangelizing their colleagues is the best OA marketing there is. Also, such a fiefdom often requires consulting departmental administration, which is a sensible opening to lobby for a mandate.
My first question is this: If faculty cannot even drag themselves to deposit material into IRs where the library has done all the tech work for them up-front, how will they be convinced to start them? It is assuredly technically simple to do, but the complexity of the technical process is not and has never been the problem. The complexity of the social process is the problem, and I fail to see how Harnad’s proposal solves it.
Indeed, if Harnad’s problem with libraries is that libraries use IRs for other purposes as well as for OA to peer-reviewed research, I honestly doubt that departmental repositories will fare any better. I can easily imagine a tech-strapped department using the repository as a content-management system for records-management and intranet functions as well as research.
My second question concerns coverage. To achieve his stated goal of 100% OA to the peer-reviewed journal literature via departmental repositories, Harnad will have to convince every department and research unit on every college and university campus everywhere containing faculty who publish in the peer-reviewed journal literature to open a repository. (He could bypass departments that exclusively publish books and monographs.) This must happen, I presume on Harnad’s behalf, without any help from librarians (since librarians are off doing their own projects) or university administration (since Harnad’s is a “bottom-up” approach). This is a tall order.
Put another way, OA proponents have yet to convince all academic libraries to start IRs, and libraries arguably have as much or more to gain as departments and individual researchers. How are departments different from libraries such that departments will universally flock to the OA banner when libraries have not?
My third question concerns unnecessary duplication of effort. Though the technical and staff requirements of setting up and maintaining a digital repository are quite small, they are not zero. What is Harnad’s justification for duplicating the necessary effort and machinery across thousands of departments? (Does anyone know the approximate number of departments and campus-based research units in the United States—just so as not to terrify myself with a worldwide number—that publish to the peer-reviewed literature? I am not sure how to make an educated guess, even.) Even small efforts loom large when so multiplied.
Still, I entirely favor Harnad pursuing this angle; there may be advantages to it I am not considering, and I stand to learn a lot about outreach to the faculty at MPOW should he succeed. More specifically, if Harnad’s approach were to make an impact at MPOW, I should think that a number of departments would say to themselves, “Yes, we want to do this, but why should we build our own repository when the Libraries are managing an IR already?” In other words, the rising tide would float my boat as well as Harnad’s; I should be insane to object.
For additional fun, let us consider some other ways OA might bypass academic libraries.
One way is state- or country-wide repositories with their own technical, administrative, and outreach staff. (I am leaving library-consortial repositories such as Aladin Research Commons and statewide-library-system repositories such as Minds@UW out, as they consist of the same academic libraries Harnad wants to cut out of the equation.) A few countries and U.S. states are trying this out, and I wish them all possible success. If this turns out to be better for OA than sharing IRs with academic libraries, excellent! we now know how to do OA, and libraries can reclaim the resources they have invested in OA for other purposes.
I myself would worry that what the government giveth, the government is liable to take away (e.g. the demise of ERIC and the current proposal to ax the EPA’s libraries), but if Harnad is right (and I am happy to grant this point) that regained research impact pays for itself many times over, a state repository should be able to make a good case for itself. If, that is, it can do a better job evangelizing or mandating OA than libraries have. I mean to reserve judgment on that until I see results from the states trying it.
Moreover, if library red tape surrounding IRs is supposed to be bad, imagine governmental red tape, and shudder!
Following Poynder, one could look to campus IT units to lead the charge. Some of these are already merged with libraries (the University Librarian at MPOW reports to the head of IT), but lay that aside for now. A campus IT unit will need to see an obvious need on campus for the project; Harnad’s “OA everywhere now!” battle-cry will not suffice, and arguments about increased research impact will fall on deaf ears because increased research impact is not IT’s concern, nor does it directly or indirectly benefit IT.
The IT unit will also need to see usage projections; they have to justify expenditures on a service in terms of how often and how intensively the service is used. In the absence of a campus mandate (in which case the IT unit would have no choice but to open an IR), usage projections are dismal and the campus-specific need is far from clear. I cannot imagine a university IT unit anywhere in this country accepting, much less initiating, an IR project under current circumstances.
Another way is additional discipline-specific repositories. I am happy to see these also, but (as I believe Harnad knows), a number of journal publishers restrict or forbid altogether deposit to disciplinary repositories, even publishers that are friendly or neutral to IRs. Until the social and legal climate changes, disciplinary repositories cannot reach Harnad’s 100% OA goal.
Finally, I can imagine not-for-profit organizations picking up the OA banner. Again, this approach faces legal and social challenges from journal publishers, but perhaps lesser ones than discipline-specific repositories; also, a not-for-profit could perhaps use its resources to fight some of these legal challenges in court and win victories for repositories everywhere. I do not really see this happening yet, but I have no objections.
My sense still is that OA needs academic libraries and academic librarians. I await further developments with interest.
