Rethinking How We Publish to Support Open Science

Abstract

With the advent of the digital age, the way we create and consume information is changing. New ways of capturing and communicating the research process digitally give us the opportunity to honour the norms of open science by showcasing and enabling the re-use of a much broader range of contributions to research. This was widely predicted when the first journals appeared online but has taken longer to materialise than many anticipated.

Research workflows have been transformed over the past few decades by the ability to gather ever growing datasets, to analyse them with ever increasing computing power, and to collaborate online. Scholarly publishing, however, has by and large lagged behind. While publications are now processed and distributed largely digitally, the publishing workflows, outputs, and fundamental concepts have largely remained artefacts of print publications.

Change is both needed and imminent. Journals have served as guarantors of quality through editorial oversight and peer review, and there is comfort in maintaining this view. However, by 20th century norms, this gatekeeping generally meant publishing only those authors and outputs with familiar credentials. This is reminiscent of those who resisted early printing presses on the grounds that they cheapened knowledge and threatened religious authority (Quocirca 2024). It is neither realistic nor desirable. Communicating research today requires expanding this view. We believe a more useful approach is to reimagine how we assess and share research, as well as how we enable discovery and reuse, while fully embracing the principles of open science.

Open science is about more than being able to read an article. It is about providing the right context to understand it, the resources to replicate the work, and the tools to collaborate and make science better. It is also about broadening participation in knowledge creation, dissemination, and reuse. We have an opportunity to make a move away from the legacy constraints of the physical format and take advantage of the opportunities provided by a digital world to support the advancement of usable, trustworthy knowledge and enable global participation.

The UNESCO Recommendation on Open Science (UNESCO 2023) “outlines a common definition, shared values, principles and standards for open science at the international level and proposes a set of actions conducive to a fair and equitable operationalization of open science for all.” Ultimately, open science is a set of principles and practices that allows science to be conducted according to its norms and “as a common good”. The UNESCO definition adds to that of the US National Academies (2018), emphasising the need for diversity of participation in order to achieve the more reliable and effective knowledge creation that open science promises.

The inception of PLOS was inspired by developments in scientific and information technology that ushered in new ways of doing science and promised a digital transformation not only of how science was done but also of how it was curated, disseminated, and discovered. We see enabling open science as the next step in this transformation.

As the international conversation started by the UNESCO recommendation exemplifies, the diversity of participants in knowledge creation and reuse is extremely important as it enables a multiplicity of perspectives which, as the historian of science Naomi Oreskes argues is the essence of trust in science (Oreskes 2021). Oreskes draws on the work of philosopher of science Longino (1990) in acknowledging that “an individual's beliefs, values and experiences necessarily affect their work—scientific or otherwise—so the best way to develop objective knowledge is to increase the diversity of knowledge-seeking communities.” Oreskes argues that science can be objective even if individual scientists are not. “Objectivity is likely to be maximized when there are recognized and robust avenues for criticism, such as peer review, when the community is open, non-defensive, and responsive to criticism, and when the community is sufficiently diverse that a broad range of views can be developed, heard, and appropriately considered.”

As highlighted by Chan (2019), it is vital to establish the mechanisms that enable this openness in a fair and inclusive way, with an awareness of the context and of who derives the benefits from openness: “…openness as a concept must be rooted in proper historical and political contexts, otherwise we risk replicating the power inequalities and asymmetry that we seek to challenge and replace.”

Today, while funding and government organisations are increasingly adopting open access and open science policies, the open science movement is also facing some serious headwinds. At least two obstacles are rooted in current publishing models.

First, publishing is still focused on articles. This is an artefact of the print paradigm which has led to the published article often being considered the only research output of value in academic reward systems. Data, code, protocols and other important outputs are not shared and valued to the same extent as articles and researchers who make them accessible are by and large not rewarded. While the proportion of articles with associated Data Availability Statements has increased, the majority of the data themselves are available only ‘on request’ or through Supporting Information files associated with articles (Colavizza et al. 2020). PLOS' Open Science Indicators show the percentage of data deposited in a repository remains below 35% (Public Library of Science 2022). This is true even for PLOS journals, which have one of the most stringent data availability policies in the industry requiring data to be made available at publication. Both the PLOS data referenced above and the ‘State of Open Data 2024: Special Report’ (Hahnel et al. 2024) also show significant regional variation in levels of data deposition. The ‘State of Open Data 2024: Special Report’ shares the outcome of a researcher survey indicating that the primary blocker to engaging with open data is lack of credit. This supports our position that for open science practices to be adopted widely, a variety of research outputs must be linked, discoverable and credited in their own right.

Second, the business models most frequently used for open access publishing are based on “fee per published unit” Article Processing Charges (APCs), although it should be noted that 80% of spend on scholarly publishing is still paying for paywalled subscriptions (Delta Think 2024). PLOS and BMC introduced APCs at their founding in the early 2000's as a mechanism to include the cost of publishing as part of the cost of doing research. The PLOS Founders noted that the transition from print to online publication meant that the “essential rationale of the pay-for-access model has disappeared” and saw an opportunity to “eliminate monopolies over essential published results, diminish profit margins and create a more efficient market for scientific publishing” (Brown et al. 2003).

While some of this change has come to pass, we also see the continued dominance of the “big deal” and the transition of open access from a driver of systemic change, to a business model supporting the status quo for many publishers. Additionally, we see the entrenchment of APCs as the dominant business model for open access. APCs are exclusionary for many researchers and institutions who cannot afford them. An economic shift is essential to facilitate access to knowledge and participation in knowledge creation by a wider range of institutions and researchers. Including those with limited resources. Non-APC-based business models, including PLOS' own Community Action Publishing and various institutionally- and community-supported models that cover costs without either reader or author fees, do exist and support born-OA publishing. But they are not easily scalable. Others, including Subscribe to Open offer a transition path for hybrid publications, but rely on transition of previous subscription revenue to support open access costs. These models by-and-large bundle the cost of assessment, publication and sharing of all outputs into the cost of publication of an article. They are not designed to consider publication of ‘non-traditional outputs’ that are not accompaniments to a traditional article.

But if publishers are part of the problem, we can also be part of the solution by grounding our models in open science principles more actively.

Although this is framed through an innovation lens, it is helpful in identifying the needs that publishing addresses—a first step towards envisioning alternative approaches.

These contributions to the research enterprise can be maintained while simultaneously broadening our horizon to an open science ecosystem. A world where publication and linking allow credit to be given for all contributions, supported by business models that are conducive to the curation, sharing, and reuse of a wide variety of outputs.

The problems we outline are systemic in nature and solving them is a long-term multi-stakeholder challenge. But at PLOS, our vision is that we can contribute essential elements of the solution and start effecting change from within publishing. We have embarked on a Research and Design project (PLOS 2024, October) to define the key elements of a new publishing model, anchored in open science principles with an accompanying business model to sustain it that can broaden access and participation.

We started the strategic development leading to this project by considering the roles and functions of publishers as outlined above, to identify which parts to keep, which to evolve, and which to “re-invent” in order to support open science. This is particularly important given the emergence of new ways to openly disseminate research: preprint servers, specialised repositories for various research outputs, journal-independent peer review, etc. We also carried out a scan of the landscape and a series of structured interviews with representatives of stakeholder groups to test and clarify our initial hypotheses (Mudditt 2022).

As much as possible, we want to work with the non-profit organisations that provide new open services and functionalities. Our intention is not to usurp their role but to integrate effectively with them and identify the changes that PLOS, as a publisher, can make to support the principles of open science more effectively.

Our ambition is to design solutions that will help catalyse systemic change and we acknowledge the complexity of the change required. To be successful, the solutions we develop must not only inspire researchers and be supported by librarians, but also integrate with the systems of the stakeholders who provide both funding and recognition. Furthermore, a diversity of perspectives is essential. Solutions to open science challenges developed for social psychologists do not necessarily translate for biologists. Interventions designed for the Global North can be counter-productive in the Global South. Teaching institutions cannot necessarily adopt the solutions designed for research-intensive universities. Therefore, a key component of this project is a broad community consultation and communications programme. We will engage with stakeholders across sectors, disciplines, regions, and institution types to test our solution designs with as wide a range of actors as possible to avoid creating or reinforcing inequities.

Our current work will extend through 2025 and in line with open science principles, we intend to share our findings as we progress through the project. We will share regular updates shaped appropriately to the content, as well as a public report towards the end of the research phase.

We hope that these will spark interest and debate, and we look forward to engaging with various stakeholders across scholarly communications as we progress.

V.K. conceptualization, writing – original draft. A.M. conceptualization, writing – reviewing and editing. N.O’C. conceptualization, writing – original draft. Authors are listed in alphabetical order.

All authors are employees of PLOS.