By Ellen Finnie

By IO blogger Ellen Finnie with Guest co-author Greg Eow, AD for Collections, MIT Libraries

As charged discussion around Elsevier’s purchase of SSRN continued in the past week, Elsevier and the University of Florida (UF) announced a pilot that links UF’s institutional repository with Elsevier’s platform. By employing an automatic deposit of metadata about Elsevier-published UF articles into UF’s repository, with pointers to Elsevier’s site for access to the articles themselves, users of the UF institutional repository will be able to discover and, if they have authorized access, access final copies of Elsevier journals.

Elsevier describes the pilot as a means of “exploring how publishers and universities can work together to make their platforms and repositories interoperable.” And in the words of Judith Russell, Dean of University Libraries at UF, “The project addresses several university needs including showcasing UF’s body of works, providing a better user experience for researchers who use its repository and facilitating compliance with US policies on public access to federally funded research.” liveporncams

While proponents of this pilot suggest a number of potential benefits, at MIT Libraries our initial take is that this approach does not align with our vision for scholarly communications and open access. In fact, when an Elsevier/CHORUS team asked us to participate in a similar pilot program, the MIT Libraries declined. Our repository [email protected], like many others, was designed to be, and has remained, an open access repository, not a de facto discovery layer for commercialized content.   The MIT faculty recognized as they developed their Open Access Policy in 2009 that publishers couldn’t be relied on for permanent access, or access at all—indeed, a main motivation of the faculty Policy was that licensing models in the digital era had left universities without assured access to scholarly articles, vulnerable to losing access to even their own research output. In ceding data gathering, reporting, and article access to a for-profit commercial entity legally bound to focus on profit for stockholders, this kind of pilot risks what we hold dear – our ability to access and build upon prior science and scholarship in the service of humanity.

At MIT Libraries, our aim is to create a world where global access to scholarship, both today and in the future, is as frictionless as possible. And this commitment goes beyond current access to research articles, to include a commitment to build, share, and preserve our cultural heritage in support of this aim. We are not convinced that our larger objectives — including digital preservation, long-term access, and access to the fruits of scholarship as democratizing knowledge and promoting social justice – are accomplished through this kind of new partnership. In fact, we are concerned that this pilot represents a Trojan Horse strategy that, rather than making Elsevier’s platform more open, serves to undermine the value and promise of our institutional repositories by turning them into little more than discovery layers for commercialized content.

Our vision for a healthy, global scholarly communications environment is different, and it is this: a community of organizations, including libraries, museums, university presses, and government agencies building a wide-ranging open infrastructure to achieve our goal of democratized access to science and scholarship, including for example the following:

  • Shared repository ecosystem
  • Unified deposit interface for all campus, government, and nonprofit repositories
  • System for aggregated and inexpensive usage data, including research analytics
  • Nonprofit campus-supported disciplinary repositories
  • Shared print collections
  • Shared print storage
  • Shared digital collections and discovery systems
  • Collaborative digital preservation
  • Top quality open access journals
  • Less expensive, open source publishing systems and services

Admirable efforts are being made in all these directions. We have arXiv, DPLA, HathiTrust, Internet Archive, ORCID, SHARE, and other important initiatives too numerous to name. But these nonprofit collaborations have taken time to surface, are still limited in scope and execution, and are taking too long to build.

As we see it, the question is no longer how proponents of open access can work together to build a shared infrastructure for the open dissemination of scholarship, but rather how can we move quickly to jump the political, cultural, organizational, and economic hurdles so these open infrastructure initiatives can move swiftly with development and wide adoption.

The two authors of this post and the colleagues we have consulted in its writing (including those listed below) represent more than a century of cumulative experience in campus scholarly communication. Our experience tells us that this pilot is a kind of collaboration that takes us down the wrong path.   In potentially offering some shorter term benefits, depending on one’s calculus, it cedes too much to a commercial entity whose values and mission are orthogonal to our own, and sets us on a road that is in opposition to realizing our deeply held values and mission.


While the above post is authored by Ellen Finnie and Greg Eow of MIT, the following members of “In the Open” endorse and support the statements made here.


Amy Buckland

William Cross

Ellen Finnie

April Hathcock

Patricia Hswe

Lisa Macklin

John Sherer

Sarah Shreeves

Kevin Smith

Claire Stewart

Shan Sutton

Ellen Finnie

Leads the MIT Libraries’ scholarly communications and collections strategy, including efforts to influence models of scholarly publishing and communication in ways that increase the impact and reach of MIT’s research and scholarship and which promote open, sustainable publishing and access models. She is a strong supporter of therapy dogs in libraries!

Comments (29)

  1. Obviously I’m biased; but I think this is spot on. Thank you for writing this.

  2. As a longtime advocate of open access, reaching back to the 1970s, I have personally committed to making my writings as accessible as possible by depositing them with Penn State’s IR. But in a test I recently did with my latest paper, published in the April 2015 issue of the Journal of Scholarly Publishing, I discovered that a search through Google Scholar did not show my article in its Green OA version to be available at Penn State’s IR but only either directly from the publisher, the University of Toronto Press, or from Project Muse. When queried, Patricia Hswe (a signatory to your statement) explained that the metadata Penn State uses is not yet designed to facilitate searches via Google Scholar. Since this is surely one of the most frequently used search tools for discovering scholarly articles, imagine how I felt? So I underline the message you are giving here, viz., that the academic community needs to ramp up its efforts to provide the mechanisms we need to enhance discoverability.

  3. Thank you, Ellen and Greg, for providing a clear and public response to the University of Florida’s announcement. I have shared your post with the entire UD Library staff.

  4. Thank you for this clear and concise statement. In particular, thanks for recognizing that sharing and unifying the infrastructure is essential. Too much of our time is taken up with implementing one or another proprietary system, only to find out that they will never meet our needs, and that just when we get comfortable with them they will no longer be supported.

  5. Bravo Ellen and Greg! A really compelling clarification, as the perfect counter to the kinds of obfuscation we’ve all seen around publishers “working together with libraries and universities” in the realm of open access.

  6. We need to keep in mind that Elsevier is strengthening their move towards greater control the ecosystem (analytics, networks, etc.). My feeling is that we need to resist this sort of “integration,” and I really appreciate you calling out how this “collaboration” does not seem to lead us in the right direction.

  7. I agree with this statement, but I want to point out that I think part of a real equitable Open Access movement needs to also respect the right of an author _not_ to choose to make their materials open until such time as they feel they’re ready. Often institutional policies aren’t granular enough in this regard or are created for the sciences and engineering, and the push to open access there relies on models that are great for STEM, but may not work for the arts and humanities. I’m thinking, for example, of someone trying to turn their monograph into a book having increased difficulty due to the practice of requiring its placement in a functionally open ETD system.

    I find often libraries are very concerned about accessibility and discoverability, and that’s only right. I just wanted to remind everyone that equitability, especially towards authors and producers of intellectual property who don’t have a lot of power in comparison to the institution, needs to be part of the conversation as well.

    1. This raises the interesting question of whether, if presses move toward an OA model of scholarly publishing, authors of dissertations will have any hope of finding publishers for their revised dissertations.

      1. I don’t know if presses will be a problem, but I can say the now-open ETDs does have a chilling effect on the ability to find a publisher to work with in publishing the revised dissertation. I can speak from experience there.

    2. Agreed in theory. In practice, I find that libraries and IRs are _very_ sensitive in particular to the publication needs of writers of theses and dissertations, and tend to build in embargoes and other tools to accommodate everyone’s needs. The concern about authors is alive and well in IRs right alongside concerns re accessibility and discoverability. Indeed, library and IR staff have done a tremendous amount of work toward handing back to authors agency that has been largely taken away via smoke and mirrors and coercion by commercial publishers.

      See also 2011 research from Ramirez et al.; the situation with regard to ETDs being published was already encouraging then, and has been only getting better since. and

      1. I think it depends on the institution. I agree that some campuses have done a very good job, but I’m not sure the progress you mention is true across the board. I don’t want to get into specifics without throwing stones as regards my own situation, and it’s definitely possible that that situation has colored my view, but I still think there’s a lot of progress that needs to be made for the humanities.

        With that qualification regarding my likely bias said, I think it’s also important to not consider every commercial publisher or academic press as an Elsevier. I think they’re often framed as a monolithic block when that’s not really true.

      2. There was a critical discussion of the Ramirez et al. study on ALA’s SCHOLCOM listserv back in September 2014. I made this point back then:

        I disagree about the perceived effect of having a dissertation openly available on the chances to get a revised version of it published. Heck, even its availability in ProQuest’s database can have an effect on sales of books based on dissertations to the academic libraries that subscribe to ProQuest.

        A study i did of books about Latin America published by Penn State University Press during the 20 years I was director showed that the ones that had been based on dissertations sold about 20% fewer copies than books not based on dissertations. Although there may be multiple explanations for this outcome, I remain convinced that one of them has to do with librarians’ reluctance to spend scarce acquisitions dollars on books some version of which they already have in their collections via ProQuest. Acquisition editors are aware of the lower sales of dissertation-based books and may therefore be reluctant to publish them.

        Some societies, like the AHA and the Medieval Academy of America have adopted policies calling on graduate programs to allow for embargoes of up to six years. Here is the MAA statement: I think they are wise to do so.

        1. We agree on embargoes being necessary. The problem you describe with dissertations seems to be applicable to all dissertations, irrespective of open access issues around them. Open access promotes citations, and is also credited for sales increases of some research monographs that aren’t based on dissertations. This is also true for some fiction:

          So many other factors are at play, of course this is complex. I was responding to the facile “librarians/institutions should…” followed by a description of what librarians and institutions have worked hard on for decades. That we have more work to do is neither a revelation nor surprising in this hugely transformative moment. The work is being done.

          1. It’s easy to consider those concerns facile when it is not your intellectual property (or indeed, your ability to find gainful employment) at stake, don’t you think?

            It’s obvious the conversation isn’t really going to be productive from here, so I’ll step out.

          2. I am truly sympathetic to your personal experience, Matthew. I also think you are extrapolating it too far, and not acknowledging the hard work literally thousands of people have been doing all around the country on behalf of researchers like you.

          3. I am in no way ignoring the work that people have done. I am saying that the work is not seen on the ground, and that I think that the concern for dissemination is necessary but that other factors should be considered. Or perhaps, if they are considered, actually appear in policies.

            Out of curiosity, however: how exactly, does required deposit in an ETD system (which is a form of open access, is it not?) promote my work if it makes it less likely to be able to revise the dissertation for the publication necessary to either gain a position or to get tenure?

          4. I fear we may be highjacking the comments, so I’ll make this my last.

            “I think that the concern for dissemination is necessary but that other factors should be considered. Or perhaps, if they are considered, actually appear in policies.”

            They are. They do. We’re not done with the work, which means — clearly — that you have gotten the short end of this stick. But they are. They do.

            “Out of curiosity, however: how exactly, does required deposit in an ETD system (which is a form of open access, is it not?)”

            It is not. Most of the time it’s correlated with open access, but it’s not by definition a form of it. Institutional repositories are also powerful tools for digital archival and preservation of cultural and intellectual heritage.

            Common exceptions to making ETDs open access are: fiction; sensitive information that could harm living people if made open; sensitive information that is prohibited by law from being made public; patents pending; publication pending or possible. Embargoes vary from six months to indefinite.

            “…promote my work if it makes it less likely to be able to revise the dissertation for the publication necessary to either gain a position or to get tenure?”

            This seems to be what happened to you, and I’m sorry it did. From what I’ve seen in the field, the benefits of open access far outweigh the drawbacks, but that’s not evenly distributed on the individual level. I believe you can still derive significant benefit from open access, the precise nature of which depends in large part on you. Ideas about how to do that can be gleaned from many writings about OA on the internet, not least on this blog, but also elsewhere.

        2. I was actually involved tangentially in the drafting of the MAA statement (someone was asked to comment, they asked me what I thought, I sent them my thoughts and research, and they passed my materials up the chain).

          I think the policies that don’t allow for up to six years also assume that people are going to immediately find a teaching position that allows them time to work on the monograph. As we all know, the market in the last few years has made this unlikely, yet institutions are not really taking this into account in formulating their policies. The message that is sent to students is that once they leave the institution it does not really care about their success. In fact, in some cases the institution tries to claim that the dissertation in the humanities is functionally a work for hire and so they have every right to utilize and mine it as they choose.

          I really understand where libraries are coming from with as far as pushing Open Access (and in fighting back against Elsevier). As I said before, though, I just want to make sure in doing so people with very little institutional power do not have what little they have stripped away in the battle.

  8. I don’t think I entirely agree with the argument that the institutional repository cannot also serve as a discovery link. More importantly, I wonder if there is a middle course that libraries should advocate (and negotiate) that could solve all problems, i.e., negotiate with Elsevier to allow the institution to keep an open copy of any article that is published by an institution’s faculty (in this case, MIT) along with the Elsevier-generated metadata. Elsevier is already benefitting from MIT’s subscription, and those of other academic library. Even if every library in the country were to load their faculty’s articles into their institutional repositories, it is doubtful that the revenue stream to Elsevier would suffer much through such open access since many discovery services or direct searches would still find and retrieve the ScienceDirect site. What would be gained would be both open access immediately and long-term preservation.

    1. We have been approached about this as well, and our conclusion about what’s being offered and our position on its desirability at the University of Michigan has been similar to MIT’s.

      Regarding “…negotiate with Elsevier to allow the institution to keep an open copy of any article that is published by an institution’s faculty…along with the Elsevier-generated metadata,” we have made efforts along these lines, but so far have not (yet? I hope it’s yet!) met with success.

  9. […] ← Previous […]

  10. […] the full blog : Beware the Trojan Horse: Elsevier’s repository pilot and our vision for IRs & Open Access | IO… […]

  11. The members of the COAPI Steering Committee are collectively making the following statement on the UF-Elsevier pilot that links UF’s institutional repository to Elsevier’s pay-walled published versions of faculty articles. Please note we are speaking on behalf of ourselves and not COAPI as a whole.

    The COAPI Steering Committee believes the UF-Elsevier model sets a dangerous precedent that undermines our commitment to open access and the fundamental purpose of institutional repositories as tools for enabling that access. By shifting the UF repository’s focus to serving as a discovery layer for pay-walled articles, the UF-Elsevier model subverts the commonly held vision of repositories as a collective mechanism through which academic libraries empower all users, regardless of their location, affiliation, or economic means, through open dissemination of scholarship. We are also concerned about ceding responsibility for repository content to a company that has proven to be antagonistic to our institutional open access policies and the principles on which they are based. For a more in-depth analysis of these issues, we highly recommend this post on the “In the Open” blog, which articulates related perspectives that we wholeheartedly endorse.


    The COAPI Steering Committee
    June 2, 2016

    Michael Boock, Oregon State University
    Ada Emmett, University of Kansas
    Anne Langley, Pennsylvania State University
    Catherine Mitchell, California Digital Library
    Jere Odell (Chair), Indiana University-Purdue University, Indianapolis
    George Porter, California Institute of Technology
    Simone Sacchi, Columbia University
    Shan Sutton, University of Arizona

  12. Many thanks to the COAPI Steering Committee for this powerful and important statement.

  13. In November 2015, COAR published a set of guidelines for assessing publishers’ repository services. These guidelines are aligned with the comments above and underscore our vision that the role of repositories is to provide open access to content. Thanks for raising awareness of this issue.

    (copied below)

    A number of journal publishers have begun to offer services to the repository community. These services may differ according to publisher, however they generally involve the automated transfer of content (metadata and/or articles) into repositories.

    COAR has adopted the following guidelines to help our members make informed decisions about such services. These guidelines are based on the underlying values and aims of the repository community, which are to provide open access to research outputs via a sustainable network of open access repositories.

    COAR recommends that repositories only enter into publisher agreements that fulfill the following conditions:

    1. Services should include the transfer of both full text and metadata into the repository, and the transfer should occur either before or at the time of publication. Metadata linking from repositories to pay-walled, full text articles does not improve their accessibility and simply serves to drive traffic to the publishers’ sites.

    2. Services should not impose embargo times of longer than 12 months (ideally, 6 months in the fields of Science, Technology, and Medicine) and existing publisher embargo periods should not be lengthened as a result of the adoption of the publishers’ repository services.

    Repositories are also encouraged to seek licenses that enable the re-use and full-text mining of content once embargo periods are over.

  14. […] “In potentially offering some shorter term benefits, depending on one’s calculus, it cedes too much to a commercial entity whose values and mission are orthogonal to our own, and sets us on a road that is in opposition to realizing our deeply held values and mission.” From the new blog IO: In the Open: Beware the Trojan Horse: Elsevier’s repository pilot and our vision for IRs & Open Access […]

  15. […] open. It also doesn’t open up Elsevier content, which remains completely closed) and in a more sceptical blog post (which describes it as turning the repository into “a de facto discovery layer”. From […]

  16. […] content. Last year’s pilot to link an institutional repository with ScienceDirect API was strongly opposed within the library scholarly communications community for its seeming abandonment o…, but for the university spearheading the pilot it was all about monitoring mandate compliance. We […]

  17. […] Finnie E and Eow G. (2017) Beware the Trojan Horse: Elsevier’s repository pilot and our vision for IRs & Open Access. In the Open. Retrieved from […]

Leave a Reply

Your email address will not be published. Required fields are marked *