This summer, a bit of news slipped quietly through my news filters — the Modern Language Association (MLA) struck an exclusive database indexing and hosting arrangement with EBSCO. Subject indexes have struggled to find a good balance with library discovery services for several years now. While I am not privy to the particulars of this agreement, I was alarmed to learn that the terms prevent MLA products, such as the well-known bibliography, will be limited to EBSCO’s discovery service (EDS).

I am sympathetic to the MLA’s motivations for such an exclusive arrangement, however, the impacts are clear and will ultimately hurt users, as pulling out of the other discovery services will undoubtedly limit visibility of MLA’s products within the library setting. In the spirit of community, I acknowledge that all players in the industry have different agendas, and we all need to balance the books. But, information providers are at risk unless we find ways to collaborate and cooperate, rather than retreat to competitive measures that ultimately diminish content discovery. I advocate for solutions that unite content and service providers with libraries around a mission rooted in efficient and effective scholarly discovery.

Simply put, exclusive indexing agreements create dams, blocking the flow of scholarly information and putting research progress at risk. Just as the data shows growing reliance on both library discovery services and abstracting & indexing (A&I) databases, it’s a shame to see this sort of barrier to scholarly workflows. Creating dead-ends and silos in this way undermines research and shores up illegal pathways to content, signaling ongoing threats to a truly comprehensive discovery experience for researchers within the academic library.

These limitations also put libraries in the untenable position of choosing to expose some resources, but not all; to promote usage of those collections prioritized by search providers, rather than the libraries paying for both their collections and the associated discovery layer. I find it disappointing to hear of library cancellations due to a conscious, strategic lack of coverage by discovery services. After nearly 10 years of advocacy for fairness by the Open Discovery Initiative, we are still struggling to strike the right balances between publisher, library, and user needs. While the discovery marketplace is both diverse and in a state of flux, I strongly believe both content and search providers should get out of the way and allow libraries to make local decisions regarding the discovery and access that best fits their needs.

Among the formal responses to this recent MLA / EBSCO announcement, such as those from ICOLC, ELUNA/IGeLU, Gale, and ProQuest, I reached out to a few librarians on the front lines of serving students and faculty with discovery and access solutions to get their impressions. I was particularly curious how non-EDS libraries were responding, and what obstacles or opportunities they saw in this exclusive deal. Please note: The responses below reflect these three individuals’ opinions and do not constitute any official or formal response from their organizations.

Response by Amira Aaron, Associate Dean for Scholarly Resources, Northeastern University Library

As the Associate Dean of the Northeastern University Library for Scholarly Resources, I have responsibility for providing streamlined access to the best possible content and metadata for our students and faculty. I also have administrative oversight of our Primo discovery system. The recent announcement about the exclusive deal involving the MLA Bibliography, both for content and metadata, is a serious and damaging development and threatens the open and effective access of scholars and researchers to content in their fields. Exclusivity does away with choice for libraries and their users. It is a dangerous trend, especially when it restricts the open sharing of metadata. This trend has become especially disturbing as it involves those content providers, such as EBSCO, who also offer their own discovery systems. I was appalled to hear one vice-president talk about “winning the discovery system war” using precisely these kind of tactics. Librarians need to renounce these exclusive arrangements loudly and make their researchers and society members aware of them as well. At Northeastern, we did compose a letter about the MLA Bibliography decision and sent it to appropriate faculty.

Libraries are aware that discovery systems are not perfect and certainly not as effective in certain disciplines as carefully indexed subject databases, but they are evolving and improving significantly. All of these discovery systems have their strong and weak points and some function best for certain types of content. My academic library switched from EDS to Primo several years ago and we have been very satisfied with this move. Although EBSCO has managed to convince some A&I Index providers that only EDS does a credible job in this arena, I believe that this rhetoric is false and misleading. Yet, we see it quoted almost verbatim by the providers who have been persuaded to restrict their content and metadata. It is imperative that direct communication between the library and index provider communities be significantly increased and that librarians be widely consulted about the best avenues to reach scholars and researchers in their institutions to prevent further decisions that threaten open scholarship. Despite assertions to the contrary, this did not happen in the case of the MLA Bibliography.

Our Primo discovery system at Northeastern is heavily used (over 1,200,000 searches in 2017). It is especially important for undergraduate information discovery but is also well-used by faculty and researchers. If metadata is not included because of exclusive deals, it will not be found and our students and faculty will not have access to that content. Although reliance on A&I indexes has been decreasing because of avenues like Google Scholar and Primo, we still see significant value in these resources and we teach them during the course of our bibliographic instruction. We also now use our discovery system to lead users to these indexes; we are not looking to replace them. Triggered by specific keywords in searches, the Resource Recommender in Primo brings to the top databases and indexes of our choosing. There are facets on the left that include all sources of search results and we include direct links to original databases and indexes on our detailed search results. If an A&I Index is not included in the results, it will not be discovered. And, frankly, the value to our institution will be lessened and we would not necessarily be inclined to renew.

My long career in this industry, both as a librarian and as a vendor, has involved much successful cooperation with all members of the scholarly communication chain, including direct competitors (e.g., SISAC, the Serials Industry Systems Advisory Committee; the development of EDI specifications, etc.). So my plea for the future is that we return to a time where vendors, publishers and libraries all worked together for the good of researchers and scholars. Competition can be healthy, or it can be extremely damaging in our industry. Exclusive deals need to be re-thought and sharing of metadata with all platforms is critical. The future of networked research, linked data technology and other major advances in scholarly communication depends on increasingly open metadata and widespread access to all content.

Response by Ken Varnum, Senior Program Manager, University of Michigan Library

I manage the University of Michigan Library’s discovery layer (Library Search) and the link resolver, and have been involved in a number of NISO projects, including the Open Discovery Initiative, over the past five years. I believe strongly in transparency in library resource access; that businesses should be generally free to set up pricing and delivery models that meet their needs, but they need to be aware of their customers’ needs and concerns as well. While “information wants to be free” is not an economic or business argument, there is an overall good to society that comes with making information as broadly available as possible, through appropriate and understood licensing agreements. Libraries, and their users, should have their choice of access points and delivery mechanisms. I think this is beneficial for information providers as well; the more licensed paths to finding high quality content, the greater the chances the “good stuff” will be found and used, rather than the “stuff” that users in a hurry will locate on the open web when they don’t have a path to access the quality content through a library or similar institution.

Decisions such as MLA’s, should they be the start of an unfortunate trend, will only diminish the value that A&I services provide libraries and their users. Aside from the clear challenge of putting important scholarly content behind an exclusive sole provider, there is an emerging challenge that we should all be aware of. As machine learning and text classification systems that can generate “aboutness” indicators grow every more adept, capable, and ubiquitous, business decisions that place even more barriers to accessing high-quality, nuanced, human-expert indexing will have an unintended consequence: further devaluing that expertise. If the typical university or college user can find sufficiently satisfactory materials through other tools, at levels of granularity that enable her to complete her research goal, what possible incentive could there be for that student to seek out a higher-quality source? While the drive for MLA’s decision may have been to exact every drop of value from human indexing while it is still better than the machine version, I fear a perverse reaction: the product will be marginalized even sooner, as technology creeps ever closer to understanding concepts, not just matching words.

Most library resource users (for better or worse) look for the easiest and most accessible appropriate item to answer their need. Barriers established between the users and the information they employ seek only add to the devaluation of the resources they do not find or cannot access. The value that A&I services have provided, and many continue to provide, is both deep and broad. It would be a shame for this important information discovery toolkit to be increasingly locked away from users.

Response by Susan McMullen, Research Services & User Engagement Librarian, Roger Williams University

The Library at RWU supports the research needs of 40+ majors in the sciences, social sciences, and humanities as well as professional programs in the areas of architecture, engineering & construction management, business, and justice studies. Traditionally, we have always viewed subject specific A&I databases as the best way to support discipline specific research. We introduce them in library instruction and encourage their use through our research support activities. However, reliance on A&I databases is not as strong as it once was. Although librarians and faculty may still favor the rich subject content and controlled vocabulary of A&I databases, student researchers, especially undergraduates, are unlikely to seek out a subject specific database unless directed by their professor. The recent reliance on discovery interfaces is having a direct impact on the usage of our A&I databases. Although it is unlikely that the richness of the search results from A&I databases can be successfully duplicated across all discovery platforms, it is important that publishers, vendors, and discovery providers work together to find some way to bring the content of A&I databases into the discovery search results.

The recent announcement by the Modern Language Association that EBSCO would be the “sole” vendor for the MLA International Bibliography came as a surprise to many of my librarian colleagues who regularly rely on this database to support students and faculty in the humanities. Because libraries rely on a competitive marketplace to make the best database purchase decisions for our end-users, it comes as a major disappointment that one of the most widely used subject discipline publishers has made the decision to grant control and access to their database content to a single vendor. It is especially disconcerting that MLA International Bibliography content will be solely discoverable via the EBSCO EDS platform. Today, libraries are relying much more heavily on their discovery platforms to discover content. As mentioned above, the average undergraduate researcher is not familiar with specific A&I databases and is much more likely to discover them and their content via the discovery platform. If the MLA International Bibliography content is not discoverable via any other discovery service other than EDS, many researchers may be unable to uncover a significant portion of the humanities literature.

Libraries must be free to choose the discovery service that best serves the needs of their constituencies and not be forced to choose a discovery service based on deals between publishers and vendors. It is extremely unfortunate that MLA has chosen to limit discovery access to EBSCO’s EDS discovery platform. As has been pointed out in the joint ELUNA/IGeLU Response Letter to MLA, it goes against the endorsed NISO’s Open Discovery Initiative (ODI). When vendors and publishers are not willing to share sufficient metadata and pull content from discovery platforms it creates a disparity in the marketplace and a sense of distrust among consumers.

From its letter to its customers, it appears MLA International Bibliography has chosen the easy path, stating that – “Increasingly we have found the time and effort to manage multiple vendors and delivery platforms detract from our focus on that development work” and that “the majority of our customers have already selected EBSCO as their bibliography platform.” Furthermore, MLA expresses dismay with the inconsistent performance of their content on most discovery services. Rather than work with all discovery systems to provide increased metadata and equal access to MLA content, MLA has simply chosen to let EBSCO manage their content and control access. In the end, will this prove to be the best path for MLA International Bibliography?

Librarians are very frustrated by the increasing lack of choice in the marketplace. The lack of cooperation and sharing of metadata between publishers and vendors in the discovery marketplace has significantly decreased the discoverability of library purchased content. Librarians and end-users are the losers in this competitive game. Rather than publishers and vendors persistently trying to tip the scorecard in their favor, it would be most helpful if competing vendors and publishers would work together more collegially.

Given the new normal of flat and decreasing budget lines, librarians are continually re-evaluating the unique content within a particular database. How much longer can we financially sustain our preference for the expert search capabilities of A&I databases when users are relying on Google Scholar and discovery platforms as their starting and ending places for research? With MLA Bibliography subscription costs rising on an average of 5% per year and A&I usage steadily declining, it seems MLA should be doing more to promote usage of its database via a much broader scope of discovery platforms rather than limiting to just one. Will A&I databases continue to remain relevant in today’s world when much of their content can be found via direct licensing of publisher content or on the open web? How will libraries respond to this type of forced manipulation by publishers and vendors?