The limits of openness: global knowledge production and its boundaries

The open access movement has been gaining momentum in recent years. The pursuit of openness is no longer confined to research publications, but also includes data, codes, software, peer review reports and so on. Open research, open science and open access appear as key terms in national research assessments and funding mandates, while the Barcelona Declaration of Open Research Information (2024) states that “[O]penness of research information must be the new norm.” Large-scale projects have been funded to examine and explore the normative and technical requirements of open infrastructure (e.g., DIAMAS, OPERAS). Topics and questions related to open access and open research occupy the programmes of conferences, workshops and webinars. In the realm of open access, article-processing charges, transformative (also referred to as transitional and read-and-publish) agreements and the alluring profit margins of some publishers have raised concerns about epistemic diversity, global justice and sustainability of knowledge production (Berger, 2021; Ma, Buggle and O’Neill, 2023; Mboa Mkoudou, 2020; Raju and Badrudeen, 2022). When it comes to the implementation of open research, there remain many challenges including infrastructure development and best practices, not to mention institutional resources and support. The virtue of openness seems unquestionable: open access democratizes knowledge by removing paywall so that anyone with an Internet connection can peruse research publications. What are the limits of openness, however? The critiques about barriers and enclosures of research information have largely been focusing on commercial, profit-driven publishers and research infrastructures (e.g. Ma, 2023; Mirowski, 2019). Conceptually, the very notion of open has seldom been contested (Christen, 2012; Neylon, 2015). In practice, there have been few, if any, discussions about national security, geopolitics and political economy that can slow and stop the flow of research information.

This article aims to examine some neglected aspects in the discourse of openness in research and scholarship, specifically, the assumption of research information as a public good. There are two points of contention: one is that private ownership and intellectual property rights can be more conducive to public goods (Cohen, 2006; Samuelson, 2006; Wagner, 2003); secondly, information as a global public good (see, e.g. Verschraegen and Schiltz, 2007) cannot truly be realised because of (supra)national borders and economic and geopolitical conflicts. To consider the limits of openness, however, is not against openness, but to reflect on and contemplate the development of the open access movement, as well as some speculations about the future of open research in the context of recent wars and research security concerns (European Council, 2024; National Science and Technology Council, 2022).

The rest of the article is organised as follows: first, the conception of open access movement is traced to understand the relationship between open access and public goods. The following sections explore the idea of research information as a public good against the backdrop of intellectual property rights and the political economy of knowledge production, before turning to a reflection of information as a global public good by discussing recent announcements and statements of open research policy and by enacting historical incidents where information flows were prohibited. The article concludes with the significance of the centre-(semi-)peripheries dynamics in the co-construction of global public goods and the limits of openness. In this article, the terms “information” and “knowledge” are used interchangeably based on the discussions about public goods and intellectual property rights in economic and legal literature.

From non-subscription print publications to online journals on the intranet and Internet, making information freely accessible has long been an ambition (Fausto, 2013; Fyfe, 2020). The World Wide Web, digital storage and high-speed connection propelled its progress and sparked the open access movement marked by the Budapest Open Access Initiative (2002), the Bethesda Statement on Open Access Publishing (2003) and the Berlin Declaration on Open Access to Knowledge in the Sciences and Humanities (2003)—collectively called the BBB Statements (Suber, 2012) in the early noughties. The BOAI (2002) states that

The statements are not only about access, but are also concerned with intellectual property rights. There is an assumption about public good—that information wants to be free, that there shall be no intellectual property rights and that there shall be no restrictions with regard to the redistribution, reuse and reproduction of scholarly publications. The ideal of open access also implies global public good (“world-wide distribution”). At the time, there were few, if at all, considerations about national borders and regulatory blockages, or potentially oppressive or exploitative post-colonial practices. These early statements express the techno-utopian and libertarian sentiments similar to that of A Declaration of the Independence of Cyberspace (Barlow, 1996) and, later, the Guerilla Open Access Manifesto (Swartz, 2011).

It should be noted that, however, the motivation for open access initiatives in non-Western regions vary. SciELO (Scientific Electronic Library Online), one of the earliest open access platforms established in Brazil in 1998, was a government initiative (Fausto, 2013; Packer et al., 2014). The platform now serves 15 countries, mostly in Latin America. In the same year, AJOL (African Journals Online) was initiated by the International Network for the Availability of Scientific Publications (INASP), a charitable organisation based in Oxford, United Kingdom. It is currently a free hosting service for peer-reviewed journals with a strong focus on agriculture, health and African studies from 31 African countries (United Nations, n.d.). These two open access initiatives have no manifestoes or declarations. The ambition, motivation and values of SciELO and AJOL in its early days of development did not seem to carry the burden of serials crisis and intellectual property rights as in Western societies, likely because most of their scholarly publications were published by university departments and learned societies and were not behind the paywall in the first place. The Internet provided the opportunity to share widely, yet making scholarly information accessible was not a radical concept but a natural step thanks to the advance of digital technologies. These developments in the (semi-)peripheries have not been widely noted and discussed until recent years when open research becomes a global priority (UNESCO, 2021).

An oft-strong argument for open access is that research is publicly funded, therefore its outputs should be public goods. What is a public good, however? Non-rivalrous and non-excludable uses are two criteria most discussed: a good is non-rivalrous when it is not diminished by consumption; a good is non-excludable when one’s uses do not exclude other uses (Stiglitz, 1999; Wagner, 2003). Wagner (2003) states that information, in its intangible form, is inherently a public good since “the cost of providing the good does not increase with consumption, and where it is generally infeasible to exclude others from consuming the good” (p. 1001). Information can be learned, manipulated, memorised and very often transformed into something new: new ideas, new expressions and new knowledge. In this sense, information is a global public good because one’s learning does not exclude others learning the same or increase the cost of its provision, for example, the learning of a mathematical theorem is non-rivalrous and non-excludable (Marginson, 2024b). However, research publications and other outputs are tangible and do not necessarily meet the criteria of public good because they can be diminished by consumption and their uses can be excludable (Kaul et al., 1999 ; Ostrom et al., 1994; Stiglitz, 1999).

The open access movement and the current proposals of open research practices are not about intangible information, but tangible information including texts, data, codes and so on in their digital or physical form. Suber (2016) argues that digital publications can overcome the non-rivalry and non-excludability criteria to become “public good,” whereas Neylon (2015) asserts that it is more appropriate to understand knowledge as a “club good” that is less, but not entirely, non-excludable. Nevertheless, the “public good” argument is concerned with making tangible information freely available in the public domain that allows reuse, redistribution and reproduction.

However, the public good argument in the discourse of open access does not consider (1) whether the incentives for rightsholders may encourage more public goods to be produced and (2) the fact that intellectual property rights do not stop intangible information from being a public good (Wagner, 2003). Consider, for example, if publications were to be open access without any restrictions and without subscription or publishing fees, how and why would publishers manage the publication process: from submission and peer review, copyediting and typesetting to marketing and distribution? How can they finance the costs of publishing? These are not only concerns of big deal publishers, but also small and non-profit publishers and learned societies. It is not surprising that publishers did not come onboard with the open access movement immediately—not until they reconfigured business models that would be sustainable, viable and, for some, hugely profitable for their operations.

A related argument for open access and open research is that research publications should be freely accessible because researchers-authors are already remunerated by public funds. Currently, researchers-authors in most institutions retain copyright of their publications such that they can sign publishing agreements, often involving the transfer of copyright or the assignment of licence to publish (exclusive, non-exclusive, Creative Commons). Most researchers-authors do not object to the idea of open access, especially when open access publications tend to have citation advantage (Bautista-Puig et al., 2020). Then why researchers-authors have not been the most vocal and strong advocates for open access? One simple, if simplistic, reason is that prestige matters in academic recognition and rewards. Established publishers have been trusted to perform the role of gatekeeping, proven by track records of quality publications—tangible information—in their portfolio. When it comes to career considerations, especially for early-career researchers, publishing in recognized channels often outweighs concerns about open access or public good. It can also be argued that information in its intangible form is a public good whether the publications—tangible information—themselves are open access (Suber, 2016; Wagner, 2003).

Further considerations regarding the public good justification in the open access movement are as follows: First, publicly funded activities do not necessarily produces public goods. In the context of research and scholarship, publicly funded research sometimes produce public bad. That is to say, well-intentioned research projects can result in public bad when their results contain errors or are misused or misinterpreted although the bad may not have been known or understood at the time of publication. Second, openly accessible contents are not necessarily public goods. Internet contents are mostly openly accessible and are not paywalled, but they are not all public goods. Some research projects can produce false or flawed results, while some Internet contents are not in any sense good (e.g. deepnudes, hate speech, conspiracy theories). Third, not all public goods are publicly funded, some are voluntary works and others privately financed. For example, Wikipedia is not publicly funded and Google is a commercial service but it is generally agreed that both generate public goods. Last, not all public goods are free to access. Public libraries in Germany and the Netherlands, for example, charge a small fee but their services are essentially considered contributive to public goods. The assumption that research information, in its tangible form, should be a global public good is open to question and examination: whether public funding in and of itself justifies all types of research outputs to be publicly accessible, and whether open access and public good always go hand in hand, especially when openness can be exploited by commercial interests.

The expression “information wants to be free” is originated in a discussion at the 1984 Hackers Conference, where Stewart Brand said (Gans, 2015):

It has been argued that information wants to be free because the cost of disseminating information is becoming indiscernible. In 2022, 500 h of videos were uploaded on YouTube every minute (Ceci, 2024). Over 28,990 COVID-19 SARS-CoV-2 preprints have been uploaded on medRxiv and bioRxic since the beginning of the Covid-19 pandemic. The number of scholarly publications has only ever been increasing since the publication of Philosophical Transactions, often regarded as the first scientific journal, launched in 1,665 (Csiszar, 2018; Fyfe et al., 2017). However, currently the majority of scholarly publications and records are not owned or held by public institutions, scholarly communities or learned societies, but commercial publishers. The top five publishers, usually referred to as “The Big Deals”, accounted for more than 50% of all papers published in 2013 (Larivière et al., 2015; see also Fyfe et al., 2017) and the trend is continuing according to the 2019 EUA Big Deals Survey Report (Stoy et al., 2019). Many publishers or subsidiaries of their parent companies also engage in data analytics business by tracking all forms of data in research activities (DFG, 2021). In other words, most scholarly publications are not “public goods” or in the public domain. Instead, they are traded in the marketplace.

Under the banner of openness, many subscription-based journals have turned hybrid to answer the call for open access in the last decade: that is, an article can be open access if the author(s) pay an article-processing charge. These charges have been increasing at rates higher than inflation even though the costs of disseminating information are supposedly getting lower. This business model reconfigures contracts and agreements with university libraries that reinforce the dominant market share of established publishers and their profit margins. At the same time, more data is being generated to produce research metrics; data analytic products are then packaged to be sold right back to universities and research institutions (Ma, 2023). So far, it seems that much research information is not getting any cheaper or a public good, while scholarly publications have generated more revenues and profits for some publishers (Haustein et al., 2024). Notwithstanding the increasing costs of openness, open access seems to have become a must-have for all researchers as “transformative agreements” are negotiated and signed in the name of public good.

In fact, the call for open access has bred a cohort of publishers who trial and exploit the gold open access model that profiteers on article-processing charges and as a result democratising information for only those who can pay. The quality control of information varies amongst these publishers: some strive to promote open access by experimenting with the gold open access model while maintaining high standards of peer review; some are deemed predatory publishers who prey on researchers-authors who are eager to pay for their manuscripts to be published. Open access is achieved either way, but whether all information is good—or a public good—can be put into question. Furthermore, the necessity of payment marginalises the voices of the financially disadvantaged. In many ways, the gold open access model exacerbates epistemic injustice, that is, what can or cannot be counted as information, as a public good. Information wants to be expensive because more information can generate more profits in the political economy of knowledge production. Data analytics products hasten the pace of knowledge production by benchmarking, monitoring, tracking and ranking researchers, institutions and countries, urging for more frequent and higher volume of publications (Ma, 2023). The cost of dissemination is getting lower, yet research information is ever more expensive.

For a very long time, there were few discussions as to how open access can work considering the predominance, significance and prestige of established publishers. The open access movement struggled to envision the co-existence of commercial and non-profit actors, while grappling with the emergence of predatory publishers who exploit open access for cash. So far, it is unclear as to whether the ideal of open access has led to a more democratic and fairer world, but it certainly has enriched some publishers (Haustein et al., 2024). Meanwhile, the preservation of scholarly publications is currently voluntary and is largely supported by not-for-profit organisations such as CLOCKSS and institutional repositories. A recent study of 7 million articles shows that digital scholarly journals are poorly preserved (Eve, 2024)—a concerning issue especially when considering vanished open access journals (Laakso et al., 2021). If scholarly publications should be public goods, then there is a need to consider preservation—whether these records are owned by for-profit companies or research communities. Open access is not immune to commercial incentives and can be exploited for profits, where public goods (information as intangible goods) are by-products, not the goal. Nevertheless, openness has become a convenient slogan for the endless accumulation of capital. Information wants to be expensive.

The UNESCO Recommendation on Open Science (2021) states that it is “committed to leaving no one behind with regard to access to science and benefits from scientific progress by ensuring that the scientific knowledge, data, methods and processes needed to respond to present and future global health and other crisis are openly available for all countries.” The recommendation suggests that information—both tangible and intangible—is a global public good, meaning there shall be no barriers that obstruct transnational information flows. Such an ideal would require international collaboration and agreements to create the standards and protocols that would benefit all countries and regions.

The development of open research infrastructure and policy, however, are often nationally or regionally bounded. When The White House (2022) first announced their open science policy, often referred to as the “Nelson Memo,” in August 2022, its objective was to make the results of taxpayer-supported research immediately available to the American public at no cost. The memo states that “[T]his policy update builds on the Biden-Harris Administration’s broader efforts to broaden the potential of the American innovation ecosystem by leveling the playing field for all American innovators, which can help ensure that the U.S. remains a world leader in science and technology” (The White House, 2022, emphasis added). Similarly, the European Open Science Cloud (EOSC), an infrastructure for hosting and processing research data, states that its ambition is “to provide European researchers, innovators, companies and citizens with a federated and open multi-disciplinary environment where they can publish, find and reuse data, tools and services for research, innovation and educational purposes” (EOSC, n.d., emphasis added). These recent announcements and statements show that openness is not without borders and boundaries and that openness is not always about global public good as suggested by the UNESCO Recommendation (2021).

National borders, international relations and geopolitical tensions have not been much discussed in the context of open access and open research, likely because openness is presumed to be global—that information is a global public good and that openness would benefit the (semi-)peripheries. In reality, openness is not always global. The Nelson memo clearly states that the policy aims to fortify the leadership of American science and technology (see also Kozlov et al., 2023), while the open access policy in the United Kingdom was driven by economic imperative (Holmwood, 2020). Openness is not altogether about public good, but oftentimes it is prioritized for competition and leadership in science and technology as well as national security concerns. Felt (2014) maintains that the idea of competition and global race are central in European research policy.

Questions concerning openness can become more delicate and difficult during wars and geopolitical conflicts. Many scientific missions and collaboration were suspended, for example, after the war in Ukraine broke out in early 2022 (see also Hinchliffe and Schonfeld, 2022; Patil and Rentetzi, 2022). Major publishers issued statements to condemn Russia’s invasion and to provide support for Ukrainian research communities (e.g., ACS, 2022; Elsevier, 2022; IOP, 2022; Sage, 2022; Taylor and Francis, 2022; Wiley, 2022). Writing about the influences of the China Initiative during the Trump Administration (Department of Justice, 2021; Gilbert, 2023), Lee and Li (2021) state that “Scientific discovery, which is fundamentally borderless, is being politically bordered” (p. 2, emphasis added), and their survey shows that about one-fifth of American scientists of Chinese descent broke ties with China as a result of the China Initiative (see also Fischer, 2023; Marginson, 2024a). The opening up and blocking of scientific information are not without precedence. For instance, Richards (1996, 1998) documented the Soviet-America library relationship and information flows from the 1920s until the Cold War era affected by geopolitical tension and national security concerns. The historical evidence puts into question whether all information can be free-flowing or a global public good (see also Zubașcu, 2024). Daniels and Krige (2022) show that the very essence of the export control regime in the United States is to control information and knowledge, determining who can have access and what are permitted to move across borders. During the Cold War, scientists were subject to loyalty tests whether or not they were engaged in classified research (Wang, 1999). While some kinds of information and their carriers—including documents, instruments and people—cannot cross borders, others have been widely disseminated to promote American values using books and libraries as instruments of cultural diplomacy (Maack, 2001; Mariano and Vårheim, 2022). Information flows can be slowed or stopped in the name of national security, or free and open in diplomatic missions, meaning that not all kinds of information are free to travel or shared across borders and the movement and mobility of documents and persons can be restrained or prohibited. In other words, not all kinds of information are permitted to be global public goods.

Global knowledge production was entwined with imperialism and colonialism in the 1800 and 1900s. The collection of flora and fauna all over the world advanced our understanding of biodiversity and nature; the collection of peoples and objects opened doors to world cultures. In a way, open access and open data existed long before the open access movement, involving information and material cultures being “open” to be extracted and exploited by imperial and colonial powers (Bayly, 2004; Petitjean et al., 1992). Like Professor Lovell in the novel Babel said about language, as quoted in the beginning of the article, information was treated as a global public good and perhaps also infinite resources: “And if we learn it, if we use it - who are we stealing from?” In colonial India, survey operations were powerful tools “to know unknown people, chart untrodden paths, and estimate local resources”, while “meteorological observations and data were important for a seafaring colonizer and an agro-based colony” (Kumar, 1992, p. 278). Krige (2022) maintains that scientific internationalism is premised on universalism with the assumption that “scientific collaboration and consensus are possible despite national allegiances, and other ideological differences between people” (p. 7) while in reality transnational knowledge flows do not exist or operate independent of national interests and priorities. Knowledge sharing can be used as a means of maintaining and expanding world markets and global power (Falcone, 2022). Western knowledge comes to bear the “authority signature” (Mohanty, 1988) and is imported and exported across borders (Pereira, 2014). The centre of knowledge production remains Western until today. In their analysis of science and Europe, Mobach and Felt (2022) points out that CERN’s slogan as “a laboratory for the world” refers to open collaborations based on “a substrate of European values and governance” (p. 401). The geopolitical asymmetry in global knowledge production has drawn attention in recent years (Krawczyk and Kulczycki, 2021; Mills, 2020; Santos, 2007; Stöckelová and Vostal, 2017), however geopolitical histories have seldom been considered in the discourse of open access and open research.

Openness is presumed to be morally good because information is understood to be a public good. The techno-utopian view assumes that there would be no bad actors or bad information, while the libertarian view disregards government rules and regulations. The moral economy of open access (Bacevic and Muellerleile, 2018) means that the negative aspects and the limits of openness can be overlooked. The assumptions that information should be free and that information is a global public good, however, are not always true, not to mention information flows across borders are not always guaranteed. The near absence of discussions about borders, boundaries, national security and other competing interests is a testimony to the enthusiasm and good faith of the advocates of open access and open research. But if information were to be a global public good, the existing and potential barriers cannot be ignored. The power dynamics between allies, enemies, regions and countries can significantly influence transnational knowledge flows as to whether certain information would be kept secret or can be a global public good while also affecting international collaborations and mobility of researchers.

Dominant epistemic cultures and existing information infrastructures frame and shape global knowledge production, determining who can be producers (e.g. researchers-authors, universities and research institutions) and what can be legitimate information (Gray, 2020; Šimukovič, 2023). When the Mundaneum was established to become the centre of world knowledge in the late 19th century (Mazower, 2012; Wright, 2014), information was to be collected at the centre and what can be counted as knowledge would become authoritative globally. Meanwhile, indigenous knowledge and cultures could be seen as outdated and underdeveloped for scientific and technological progress, as hierarchies of cultures, ethnicities and skin colours were created (see, e.g., Downs, 2021; Poskett, 2019). The “centres” of knowledge have had cumulative advantages in documentation practices, academic publishing, scientific research and technological innovation. These advances must be examined historically to understand the trajectories of knowledge production in different regions. The imbalances of power–economic, military and political–cannot be reversed to correct global knowledge production, but the more important question, perhaps, is how global knowledge production can be reimagined and balanced through open access, open research and similar initiatives. An important step would be the decentralization of power by de-emphasising the significance of Western indexing services and publications, that is, what can be counted as knowledge—as a global public good—in the first instance.

The BBB statements were essentially Western with few considerations about the consequences to global knowledge production. It was very likely an oversight, but the building of Eurocentric infrastructure can resemble colonial practices. Even if unintentional, these practices reinforce and reproduce the authority and supremacy of Western knowledge: on the one hand, open access strengthens and accelerates knowledge production at the centre; on the other hand, open access serves as hegemonic power as “the rest of the world” will read and depend heavily on research from the Western regions, but not vice versa. Together, these two effects can have significant impacts on the power imbalances of global knowledge production. Further, national priorities mean that open research mandates would be partly directed based on national interests, whether it is about research security, international competitiveness or world peace.

Furthermore, information as a public good is contingent on the fact that information—tangible or intangible—that there shall be no bad information or disinformation. However, the history of science has shown many precedents where knowledge production can be used to misguide and misinform (Oreskes and Conway, 2010; O’Connor and Weatherall, 2019), not to mention when academic publications are commodities, information can be very expensive and profitable as evident by the rising costs of subscription, article-processing charges and profit-margins of a handful of publishers. It is because there is a seemingly unlimited demand for research information and an assumption that the more information, the better we are informed. But the ever increasing demand for information, incentivised by the academic reward systems means that profit-maximizing publishers are driven to increase the volume of publications, leading to an amalgam of issues involving peer review crisis, replication crisis and research integrity (Biagioli and Lippman, 2020; van Noorden, 2023) that can negatively affect the quality of scholarly publications. With these ongoing challenges, not every bit and byte should be counted as information or we risk foregoing information as a public good.

Open access existed long before online access. Distribution was limited because of the physical forms of publications as well as social and political structure (Fyfe, 2020). The availability of mass education, public transport, public libraries, community centres influence what kinds of information can be accessed and who can benefit from and further generate public goods. Paywall is one but many barriers to openness. The public good justification makes a good cause for open access and open research. However, the assumption of (research) information as a public good can be appropriated for commercial interests, national priorities, other agendas or all of the above. Tangible information is mostly private properties, while information flows are bounded by laws and regulations and can be obstructed by borders and boundaries.

Challenging the assumptions of openness is not to discount its importance or benefits, but to initiate a conversation about the political economy of knowledge production, on the one hand, and the influences of regionalism and internationalism, on the other. Both can be seen as obstacles to open access, but neither will go away. The global public good justification for open access demands further consideration: whether open access always leads to public good and whether the cost of open access is justifiable (Haustein et al., 2024). As diamond open access (i.e. no subscription fee and no article-processing charge) gains traction, it is essential to consider the mechanisms by which journals and publishers, old and new, will be supported: how to maintain quality control and whether there can be potential control and censorship under authoritarian regimes or research security concerns (European Council, 2024; National Science and Technology Council, 2022). Most importantly, history and global politics (Eve and Gray, 2020), as well as socio-economic and legal structures should be examined to understand the boundaries and limits that hinder or foster transnational information flows in global knowledge production.

The author thanks the anonymous reviewers for their invaluable insights and suggestions.