Towards a Cartography of Trust in Knowledge Production

Willa Liburd Tavernier (she/her) is the Research Impact & Open Scholarship Librarian at Indiana University, Bloomington. She manages open scholarship resources at IU Bloomington Libraries and provides publication data and data analysis to library administration, as well as colleges and departments, for institutional decision-making. Willa is committed to advancing inclusion and belonging in her work and research. Her research interests are equitable scholarly communication, governance and sustainability.

Why should we pay attention to trust in debates on open access? In 2023 Clarivate’s Web of Science delisted 50 journals citing integrity concerns in a move described as ‘just the tip of the iceberg’ while reviewing 450 more. Nearly half of those delisted were open-access (OA) journals. Criticism highlighted high-volume special-issue throughput managed by guest editors with questionable expertise, and the fee-based business model of the publishers of these journals via article processing charges (APCs) also known as gold OA, a model that can incentivize raising acceptance rates. Delisting or potential delisting of open access megajournals is currently in the spotlight.

In the 22 years since the Budapest Open Access Declaration, academic libraries shifted their missions to support open access, hoping it would reduce costs and democratize knowledge production. Currently, the dominant model is gold OA which is supported by libraries and their institutions through open access funds and transformational agreements, while authors and institutions without the financial means to pay them are dependent on the largesse of publishers who offer APC waivers. This has exacerbated financial strains rather than alleviating them, as libraries now face higher demands to support OA publishing while still spending significant amounts on access via subscription agreements.

Over the same period, public mistrust in scholarly communication has grown, fueled by misinformation and the erosion of traditional research safeguards, especially evident during the COVID-19 pandemic. Issues like mandatory publishing quotas, a lack of editorial freedom, and deceptive publishing practices have further strained the system. Layering OA on top of the existing commercialized scholarly infrastructure has not yielded the anticipated benefits of democratization and reduced costs. There is also a long history of mistrust of scholarly, and, in particular, scientific information within vulnerable communities which have been abused and targeted by governmental authorities, scientific and medical professionals, researchers, and journalists, so mistrust of research or of scholarly knowledge is not a new phenomenon nor is it necessarily a consequence of open access. However, trust issues affect the conception of open access as an unmitigated public good. 

For libraries the thinking has traditionally been that the problem is access, and OA was seen as a tool to expand access. Richard Jefferson made the point earlier this year that access is not enough to democratize the benefits of scientific research and translate it into beneficial public outcomes. He noted that useful knowledge and supporting infrastructure is what is needed. However, knowledge is only useful if it is trusted – both by the scholarly community and the public. Useful, trusted knowledge is the resource on which the scholarly communication ecosystem relies. Therefore, along with expanding access to knowledge, the stakeholders in the scholarly communication ecosystem – researchers and their institutions, publishers, indexers, funders, government agencies, scholarly societies, and libraries – must also ensure that trust is preserved and expanded rather than diminished. The trustworthiness of knowledge is diminished by manipulation or misuse, for example through disinformation, misinformation, falsification, and misrepresentation, all of which have played out in the public sphere as noted above. 

The public sphere is not the only point at which trust in scholarly information has been damaged. I recently sat down with members of the research community in Sweden for a study I’m doing on community driven publishing, knowledge production and the public good. I focused on Sweden because a very high proportion (over 80%) of Swedish research is published open access in comparison to similar income countries. In addition, Sweden was an early mover in the transformative agreement space with a Frontiers deal signed in 2018 and a more expansive agreement signed in 2024. The Swedish Elsevier cancellation in 2018 was eventually replaced with a transformative agreement spanning 2020-2027, along with similar agreements with SpringerNature, Taylor & Francis, and Oxford University Press. Trust was one of the many topics we spoke about during the study, as I was curious about the interplay between open access and the production of trusted useful knowledge. The individuals I spoke with, equal numbers of men and women participants comprising researchers, librarians, policymakers, indexers and publishers, spoke to trust vulnerabilities as a systemic issue while acknowledging the trust vulnerabilities introduced by the APC based OA publishing model, reinforcing the message that access is not enough to democratize the benefits of scholarly production. 

They identified various points at which trust has been damaged in the scholarly communication ecosystem showing that libraries and researchers are well aware of the issues. Two messages that came out strongly from these participants in my research are a weakening of the sense that the different stakeholders in the ecosystem share the same values and are working towards the same goals, and a weakening of the sense that we are able resolve issues around access to research as a community.  The process provided from the research community itself a cartography (or map) identifying points where trust vulnerabilities in scholarly communication exist. Librarians and researchers are already keenly aware of these issues, and are creating solutions. These are shown in the infographic below. Although the issues identified can only be taken as representative rather than exhaustive given the small size and relative homogeneity of the participants, I found the interactions instructive and I hope others will too.

This mapping exercise shows that the scholarly communication ecosystem primarily incentivizes the quantity of publications and citations – the publish-or-perish paradigm that the research community has known for decades. It also shows how this impacts trust in a dynamic scholarly landscape with shifting modes of and models for knowledge production and dissemination – APC based OA publishers must balance the very real priorities of quality vs. financial stability which is driven by quantity under that model. Commercial publishers have developed the infrastructure that supports these quantity based paradigms, and institutions and libraries support this infrastructure financially. 

The alternative is to incentivize trust in scholarly communication. Why should this be a concern for libraries? Libraries help incentivize both paradigms through subscription and transformative agreements. If libraries are invested in ensuring a knowledge production system that is founded on useful, trusted, knowledge, then libraries should be invested in incentivizing trust as well. 

The participants gave some examples of efforts underway at various points of the scholarly production system to incentivize trust, and there are many other efforts underway. However, none of these efforts have seen the near universal adoption that the publish or perish system, which incentivizes output, currently enjoys. These included:

1. Incorporating badging systems for compliance with indigenous data sovereignty requirements, and for implementation of responsible AI policies in publishing into discovery platforms.
2. Developing computational processes for verifying reproducibility of results (reproducibility reports). 
3. Expanding registered reports i.e. publishing protocols prior to research.
4. pre and post publication peer review, and open peer review.
5. versioning of articles allowing for identification and correction of errors even after publication, bolstering reliability.
6. Detailed contributor records for transparency and accountability.
7. Journal startups to publish research for which there were no appropriate commercial publishing options.

The landscape is ripe for libraries to partner with and support researcher driven efforts and society based publishers to incentivize trust, in the same way that this is happening in the commercial space. For example, DataSeer technologies recently announced a partnership with AAAS (American Academy for the Advancement of Science) to pilot automating MDAR (Materials Design Analysis Reporting) completion as a pilot to test the feasibility of expanding MDAR to disciplines outside of the life sciences. The standardized reporting framework enhances transparency and reliability. Libraries can look at some of the innovations underway and provide financial support to scale up these efforts, but to do so they will need to make hard decisions about redirecting their limited funds.

Indeed, on the library and consortia side, participants noted that while Transformational Agreements starve library budgets of resources for investing in other areas, just like subscription Big Deals have done in the past. They suggested that libraries should carve out fixed percentages of the amounts spent on subscription and transformational agreements to support initiatives designed to support trust in scholarly communication. 

Library investment would be a step in the right direction, but more is needed. While I was still in library school a senior librarian told me bluntly that libraries will not be able to accomplish a shift to open access unless the system for academic advancement changes (also referred to as the merit system or tenure system). The same applies to this issue. Even if libraries invest in mechanisms to address trust vulnerabilities in the system, this will not be sufficient unless institutions and funders divest from publications and citations as research currency, to a more holistic, balanced, quite possibly slower, and certainly more thoughtful approach to scholarly evaluation and assessment. As one participant put it, “scientists are publishing for the advancement of scientific careers, instead of for the advancement of science.” This puts research assessment and evaluation practices center stage.

The Declaration on Research Assessment (DORA), another initiative from the scholarly community, is a recognition that the current model of research assessment needs to change. DORA provides tools and case studies for alternative means of evaluating research, and using metrics responsibly. Again, broader, systemic adoption of these practices is needed, and librarians and researchers can take action within their institutions to raise awareness of DORA. 

Librarians and researchers cannot do this alone. Participants in the study pointed out that none of this can be the work of a few brave researchers. It must be supported by their disciplinary communities. Researchers at any stage can conceptualize and propose new modes of publication and assessment, but senior figures, in particular, will need to lend their gravitas to these worldbuilding efforts. It requires the efforts of entire scholarly communities to intentionally imagine, design, and financially support scholarly communication futures founded on trust. 

It also forces us to deeply engage with the fact that the very nature of quantity based publishing paradigms de-incentivize trust. As such we should be looking at pathways for paradigm shift. The Global South is leading the way with networks like Scielo and Open Research Africa, while the DIAMAS project seeks to build a “aligned, high-quality, and sustainable institutional OA scholarly publication ecosystem” for the Europe Research Area in the future. 

A useful exercise to prompt reflection within libraries may be to compare the level of funding going toward library, scholar-led and diamond open access publishing, with the level of funding going toward commercial gold-OA publishing. Another would be for a library to develop a values statement for decision making in this space and compare how well the publishers with whom it enters into or contemplates entering into an agreement aligns with these values. The work of the Library Partnership Rating Rubric is very useful in this regard, and there are publishers who have volunteered to be rated. 

Charting the ways in which publishing, trust, and open access intersect in the scholarly communication landscape provides a lens through which stakeholders can continue to question the entire paradigm of knowledge production, support efforts already underway to improve the scholarly production ecosystem, and open the door to a future where scholarly communication serves as a reliable and inclusive pathway for knowledge.