It has been interesting to observe the history of scholarly publishing over the last couple of decades, especially the efforts to make content available through OA channels. Despite great efforts, subscription-based business models with paywalls blocking researchers who are not affiliated with subscribing institutions prevail. Today, OA still represents a minority of newly published scholarly articles, although it does seem that recent events may have moved the industry to the tipping point for new business models.
Early Optimism for Digital Libraries and Open Content
Turning back the clock a couple of decades, I recall a great optimism that the transition from print to electronic journals would bring rapid and inexpensive publishing of scholarly articles and that libraries would play a central role. It was thought that research papers would be easily published on library-managed websites that could be operated at a tiny fraction of the cost of print subscriptions. The vision of all-digital libraries was often suggested as a near-term possibility. I attended one talk in which a library director suggested that any staffers dealing with journals would end up doing other kinds of work in the library since managing the new electronic versions would be so easy.
What has transpired is a wholesale transformation from the print publishing of scholarlyjournals to electronic formats. This conversion has been comprehensive for almost all newly published materials and, in most cases, backfiles of journals volumes issued in print. Libraries have been able to repurpose considerable amounts of physical space dedicated to runs of print journals, often to new types of spaces for study and collaboration.
The optimistic views for scholarly publishing based on open content and inexpensive access were not realized. Instead, commercial subscription-based publishing continues to prevail. The annual increases in subscription prices have pressed library budgets into difficult straits. Libraries must make difficult decisions on what titles to cut each year and, in many cases, spending on books has been slashed. The work involved with managing electronic resources has proven to be even more complex than that of print journals. The complexity of managing electronic resources has driven the development of specialized tools and knowledgebases and is a key factor behind the development of the new generation of library services platforms.
Scholarly Publishers Transformed and Strengthened
The scholarly publishing industry has proven to be resilient and was able to navigate the transition from print to digital formats. Through ongoing consolidation, the primary commercial publishers of scholarly content have become larger and more powerful over the last 2 decades. In recent years, many of these companies have expanded into analytics and workflow tools and have gained involvement in a broader scope of research and publication processes. Only in limited circumstances have libraries become the primary publishers of scholarly content, but they continue their role as intermediaries between the scholarly publishing ecosystem and the consumers of that content.
Many libraries have entered into contracts with some of the major publishers that include access to very broad packages of content for a negotiated subscription fee. These big-deal agreements can result in gaining access to substantially more content than selecting ejournal titles individually, but they tie up large portions of a library's budget. These deals include content spanning a wide range of disciplines and may involve materials not within the scope of the institution that go unused. Such arrangements can be attractive when considering cost per title, but libraries continually reassess their overall strategic value relative to more granular content selection.
Institutional Repositories: Capture the Intellectual Output of the University
Another thread of activity in the history of OA scholarly content involves institutional repositories. Based on open source software (such as DSpace and Fedora), institutional repositories were considered an important part of the ecosystem of access to scholarly papers, providing additional points of access. Researchers would place preprints of articles into the institutional repository, providing early access to research as formal papers worked their way through the peer review and publication process. Final versions could also be deposited, depending on the copyright arrangements with the publisher.
The vision of the institutional repository has been partially realized, with uneven success among institutions. It is a matter of opinion whether institutional repositories have had transformational impact on the scholarly publishing industry. Despite considerable efforts from librarians, not that many universities have been successful in fulfilling the ideal mission of the institutional repository as comprehensively capturing the intellectual output of the institution. Some universities have mandated that faculty members participate, although it is more common for the deposit of research articles to be optional. I observe many cases in which the institutional repository houses student papers, university publications, and a selection of research papers, but has not achieved a position as a primary representation of the research conducted at the university.
OAI-PMH: Interoperability for Institutional Repositories
Another key concern relates to how materials in institutional repositories can be discovered and accessed by researchers. Researchers would not necessarily want to visit each institutional repository to find articles of interest, but they would rather depend on broader search services, possibly organized by discipline, that would simplify access and discovery. The need to extract the metadata from all the institutional repositories resulted in the creation of the Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH), which remains in use today. The Open Archives Initiative began in 1999 in a meeting called the Santa Fe Convention, which articulated the need for such a protocol. This notion was refined and developed in subsequent meetings. I had the opportunity to attend the 2000 meeting of the Open Archives Initiative wherein the early concept of OAI-PMH was drafted. The basic idea involves an initial bulk harvesting of a repository's metadata, with subsequent incremental updates of added or changed items. The protocol does not harvest the documents, but rather provides metadata so that search services can enable users to access articles from the original repository. OAI-PMH has turned out to be a useful mechanism for harvesting metadata | from all types of content repositories. It I has been widely implemented, and it's one of the mechanisms used for tasks such as making the content of repositories available through general index-based discovery services (such as Primo, Summon, WorldCat Discovery Service, and EBSCO Discovery Service).
OA Gaining Momentum
In recent years, the scholarly communications arena has seen considerable movement toward OA publishing. OA publishing implies new business models other than the subscriptions and paywalls. The most common arrangement involves an advance payment of an article-processing charge (APC) once it has been accepted for publication. Some journals are fully OA, entirely supported by APCs. Many take a hybrid approach, with a mixture of OA articles and others that are restricted to subscribers.
The scholarly publishing community has taken a gradual approach to OA. Subscription-based licensing continues to dominate, although the pressure for publishing articles as OA has spiked in recent years. Many universities and grant-making agencies now mandate that any work published based on the projects they fund be made available as OA. Examples of the funding agencies requiring OA publishing include the National Institutes of Health (U.S.), Wellcome Trust, and Research Councils UK. Universities in the U.S. with OA mandates include Harvard University and the Massachusetts Institute of Technology (MIT), and there are many others internationally.
OA has seen intense movement recently through a variety of large-scale initiatives. One example is Plan S, which came out of the cOAlition S (a group of influential European research agencies). Its agenda makes a bold demand: "By 2020 scientific publications that result from research funded by public grants provided by participating national and European research councils and funding bodies, must be published in compliant Open Access Journals or on compliant Open Access Platforms." Plan S includes among its principles that hybrid OA journals do not meet its expectations and that research must be made available in pure OA publications.
Plan S has seen mixed reactions among research institutions and the scholarly publications community. Some publishers push back against some of the precepts of the initiative as not sustainable. Universities and funding agencies seem generally favorable. This more aggressive strategy comes at a time when many have become impatient with the slow, incremental approach toward OA publishing.
One of the emerging business models demanded by national initiatives and other large consortia is that of publish-and-read. Agreements based on it require that all material published by researchers within the jurisdiction of the organization be made available to the world as OA and that the publisher's entire portfolio be made available for access at a similar level of cost as incumbent big-deal subscriptions. An example of this business model can be seen in the recent agreement between Project Deal, a nationwide consortium in Germany, and Wiley (projekt-deal.de/about-deal).
The insistence on wholesale availability of scientific research through OA can also be seen in multiple national consortia. Several countries, including Norway and Sweden, have canceled their contracts with Elsevier as well as with important research agencies such as the Max Planck Society. The University of California made headlines this February when negotiations for a read-and-publish deal with Elsevier broke down and a decision was announced that it would not be renewing its big-deal contract.
Impact on Library Technology Infrastructure
These items are just a few snippets from the events currently playing out in the scholarly publishing arena. It seems increasingly likely that there will be a seismic shift in current business models. Possibilities include read-and-publish deals for large-scale organizations and increased reliance on APCs instead of library subscriptions. It does not seem especially likely that any one model will prevail universally, resulting in even more complex requirements for libraries charged with managing institutional access to scholarly materials. The eventual transition to these new models will have implications for several types of systems used or deployed by the library.
Library services platforms and standalone electronic resource management systems will need to incorporate a new set of business rules. These products were initially focused on the management of subscriptions. A library would have to acquire, renew, or cancel subscriptions for content packages selected to meet the research needs of the institution. The products used to manage a library's collection of electronic resources now have to also support OA materials. This support will, at a minimum, include tracking and analysis of the use of these materials. Many libraries may also be responsible for paying APCs for articles published in their institutions, implying yet another layer of business processes to track and manage related budgets and payments.
The discovery and linking services provided by libraries are also tied to these business models. In most cases, a discovery service would be configured to search only within the materials to which the library has provided access through its subscriptions. Articles outside that body of subscriptions could be purchased individually or acquired through document delivery or interlibrary loan services.
The advent of OA means that these resource management and discovery services have to take into consideration the body of materials available to all, regardless of whether the library has a subscription to the product. Tracking articles available through OA can be difficult, given the lack of mechanisms to systematically identify them. The NISO Open Access Metadata & Indicators recommended practice addresses this need, but it is not yet universally implemented.
Several browser plug-ins have been developed that help address this issue. Products such as the Open Access Button, Kopernio, and Unpaywall aim to lead individuals to OA versions of articles they find in search results. These services may use the Unpaywall database of OA articles or other methods to automatically link to the PDF of an article from a citation viewed in the browser. These products have caught the attention of the industry with the acquisition of Kopernio by Clarivate and Lean Library by SAGE.
The possible wholesale transition of the scholarly publishing ecosystem to OAmay also have implications for the institutional repository. Once the primary copies of research articles are available as OA from the publishers, the need for institutional repositories maybe diminished. Some institutions will continue to require representation of their research within their local infrastructure. Others may find it less important to channel efforts maintaining institutional repositories once OA has become the default method for publishing.
Following a period of at least a couple of decades in which OA publishing has been an unrealized ideal, we may finally be on the cusp of significant progress. It seems like the scholarly publishing arena has reached a point of no return to established models. Many powerful organizations that control the funding of the research are demanding change. Furthermore, Sci-Hub has wreaked havoc in another way. The availability of most scholarly articles through this illicit source has also dented the armor of the publishing firms.
Libraries need to be prepared for a new cycle of change. We need to ensure that the technology systems we employ will support the new business models that may prevail. Should OA eventually dominate the ecosystem of scholarly communications, libraries may need to adjust many aspects related to management, discovery, and access of these resources. The environment is not likely to change overnight, but now is the time to start considering how library technology products and the underlying business processes will need to be reimagined to meet the realities of the upcoming waves of change in the scholarly communications arena.