OCLC-LIBER Open Science Discussion on Scholarly Publishing

Recently OCLC Research and LIBER (the Association of European Research Libraries) hosted the first of seven small group discussions comprising the OCLC-LIBER Open Science Discussion Series. This discussion series, which takes place from 24 September through 5 November 2020, is based upon the LIBER Open Science Roadmap, and will help guide research libraries in envisioning the support infrastructure for Open Science (OS) and their roles at local, national, and global levels. I wrote about this collaborative effort between our two organizations in an earlier blog on 28 September.

Our first small group discussion focused on the topic of Scholarly Publishing, included participants from eight countries and two continents, and was facilitated by Rachel Frick, Executive Director of the OCLC Research Library Partnership. The goal of this and all focused discussions in this series is to:

  • Imagine and articulate what the ideal open science ecosystem will look like 
  • Identify barriers toward that future
  • Envision how the library community—working together—could take collective action, in order to address the challenge and effect change

The overarching goal of the discussion series is to inform our organizations as we seek to identify research questions that OCLC and LIBER can collaboratively address to advance Open Science. 

What does the ideal future state look like for scholarly publishing?

Participants shared their goals for an open and accessible scholarly publishing ecosystem, which they believed should have the following characteristics: 

  • The open scholarly publishing system will be flexible enough to support a diverse and dynamic landscape of publishing options–facilitating innovation, sharing, and discovery.
  • This new ecosystem must be sensitive of, and adaptable to disciplinary differences. Participants remarked on how divergent disciplinary norms can be in publishing (for instance, practices around author name order). And data management and sharing practices also vary widely.
  • Published digital content will not only be open, but it will also be systematically preserved. This in turn will help inspire confidence in researchers. Participants agreed that we need a landscape where scholars can be confident in the preservation of digital, open content. Humanities scholars are sometimes concerned about this, and it slows their participation in open publishing. 
  • Like research data, scholarly publications should be FAIR (findable, accessible, interoperable, and reusable), and discussants emphasized that metadata should be standardized, interoperable, and machine readable—a reality that is far from realization today, particularly for green open access existing in repositories. 
  • Persistent identifiers like ORCIDs and DOIs should be used across all disciplines and for all types of content, including publications and datasets, but also grants, proposals, and more. Relatedly, participants expressed concerns about the uneven adoption of ORCIDs and DOIs for much book scholarship, in great part because of the failure to implement PIDs in book publishing workflows. This makes for less robust information for discovery and access, and also hinders the potential for metadata harvesting of humanities content into CRIS/RIM systems.
  • Open Science terminology should be free from jargon and easy for scholars of all disciplines to understand and value. One participant in our discussion, who joined a university library after a career operating on another part of campus, expressed concern about the amount of jargon that may impede rather than accelerate the understanding, acceptance, and adoption of open science practices.
  • Researchers and scholars should be incentivized to publish open access. Research universities are heavily attuned to research metrics, and our participants suggested that open science metrics–like the number or percentage of open publications produced by an institution–are also vital for charting an open future.

What are the main challenges and obstacles preventing progress toward this ideal state?

Once we had a sense of what the ideal open publishing destination looked like, we used online polling to brainstorm a list of barriers–or roadblocks–toward progress. Participants offered about 20 items, although some are closely overlapping. Using up- and down-voting, we quickly identified three top challenges that we spent more time discussing:

  1. Evaluation and funding mechanisms for institution and researchers
  2. No agreement around which standards to back–let’s all get behind CrossRef, DataCite, ORCID, etc. and stop trying to develop different solutions
  3. Many researchers are still not aware about Open Science practices

How can the library community, working together, take collective action—in order to address the challenge and effect change?

Show me the money

Several discussants described the need for scholarly publishing and access to be an institutional priority and not just a library priority. Currently support sits primarily in the library budget for subscriptions and licenses, but this support could also logically reside in the research office budget, as they are research support expenditures. One participant said, 

“In terms of funding, it would behoove us to think about our collections budgets in terms of monies that support the actual research process. Things are going to have to change internally as to how scholarly communications is funded.” 

Participants recognized the challenge of moving open science awareness and support out of the library and into other campus units. Collaboration and partnership with other university stakeholders is only going to grow in importance. The research office is seen as a primary stakeholder and partner in this regard, and, as one participant described it “most researchers are [already] attached to the research office, whereas they may not have relationships with the library.” However, localized dynamics mean this will look different at different locales. One participant described their experience working on two different university campuses—one with a strong office of research and another with a weaker, less centralized research office. Advocacy and change is easier in the former environment. Note that OCLC Research has recently published the research report, Social Interoperability in Research Support: Cross-Campus Partnerships and the University Research Enterprise, of which I am a co-author, on this topic of intra-institutional collaboration and the need for libraries to effectively engage with other campus stakeholders. 

One participant described the shift to open scholarly publishing going in one of three ways: “legal, flipping, or revolutionary.” This would happen through 1) transformative agreements, where the content is legally shifted from subscription to OA, 2) flipping from the prioritization of gold open access to the acceptance of green open access, with shorter embargo periods—and emphasizing that green OA is not less effective or important than other methods, and 3) starting more university presses to change the paradigm, with the more radical or revolutionary change of funneling money to smaller publishers instead of large behemoths. 

Show me the metadata

Participants applauded the use of open, persistent identifiers like ORCIDs, DOIs, and RORs (Research Organization Registry), and see them as essential for supporting disambiguation and interoperability. However, they pointed to the uneven adoption of PIDs across the scholarly communications landscape and one participant recommended the adoption of ORCID for “everything and anything,” including grants, research proposals, publishing, repository deposits, and much more. 

Photo by Alain Pham on Unsplash

Another discussant emphasized how imperative it is for us to integrate these identifiers into all our systems, as this lack of infrastructure inhibits interoperability, and later, discoverability and access. Dspace lacks a native space for making ORCIDs usable and discoverable, for instance, and publishing workflows, particularly for humanities monographs, usually don’t capture ORCIDs, and might not even be friendly with DOIs. (And forget about being ready to ROR!) While many STEM publishers have been requiring ORCIDs in their publishing workflows for a few years now, large disciplinary gaps remain, particularly in humanities publishing. These gaps flow downstream, as library discovery systems may fail to pick up the 856 field that indicates if content is open access, meaning that open content exists but can’t be discovered or accessed by the user. 

The group also discussed how these failures exist because of a lack of social interoperability (my term) between different silos in the research life cycle. Some discussants recommended greater involvement from the research office. And that librarians needed to better understand the publishing workflow in order to advocate for change—and to integrate better into library services. There are also opportunities for improved collaboration between scholarly communications librarians and technical services librarians. And of course, ILS service providers must also recognize the importance of collecting and making actionable information about open access availability. 

Show me why I should care

In addition to metadata, conversations also seem to come back to issues of communications. One discussant described their weariness in this regard, 

“Sometimes I’m really tired of the communications, because it’s just broadcasting. But this is not effective enough. A kind of tailor-made communication and interaction and somehow being more visible and a part of the researcher process is essential.”

Photo by Jon Tyson on Unsplash

This rang true with others, who recognized that broadcasting doesn’t work because of the significant disciplinary differences—while terabytes of data matter to one community, it’s a non-issue for another and will fall upon deaf ears. Instead, discussants agreed that effective open science communications “will mean myriad things,” and must be customized and discipline-based. One non-library discussant also warned against the use of jargon in our communications, as library-speak may not jibe with terminology well known by researchers, especially across all disciplines. Efforts like the recent OAPEN OA Books Toolkit, offer resources to help raise awareness about open access books. And posters like this one from the ARMA conference offer a way to educate researchers about why they should care about PIDs. 

I found this a rich, positive conversation, and one that OCLC and LIBER will be collectively reflecting upon in the coming weeks, as we also accumulate input from the community on the other six focus areas in the LIBER Open Science Roadmap. Stay tuned for more blog posts in this series.