120 years after the Budapest1, Bethesda2 and Berlin3 declarations (the B Declarations) open access seems to be no longer a matter of activism, controversial, marginal or even political. Open access reached its tipping point several years ago and entered a new stage of development as a mainstream practice, almost the default way of publishing research results. As a matter of introduction, I propose to take a look at these dynamics after and maybe beyond the tipping point.

2In 2013, the European Commission published a press release4 presenting the results of three studies by Science Metrix5 showing that in Europe at least more than 50% of the articles published two years before were open access. 10 years after having reached that tipping point it could be useful to assess where we stand and what happened to the dynamics that were identified.

3A few days ago, the French Ministry of Research published a new version of its Open Science Monitor6 based on diverse sources, such as Unpaywall7, the Directory of Open Access Journals (DOAJ)8 and various repositories. I encourage you to have a look at this interesting instrument developed by Eric Jeangirard and his team with open source technology, open data and a lot of results as food for thought. In particular, the monitor evidences the strong global dynamic of OA adoption across the whole French research community, with 62% of the papers authored by at least one researcher affiliated with a French institution published open access. Of course, OA adoption is different from one discipline to another, and the strength of this adoption can be different, but it is clearly present and high.

Fig. 4: 62% and counting in France

4The same dynamic can be observed in other European countries, such as in Germany thanks to the Open Access Monitor9 funded by the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG10). The data provided by the German OAM are not comparable of course with the French Open Science Monitor because they do not proceed from the same methodology. However, for example, we have data showing a proportion of 61.4% of articles authored by researchers affiliated with a German institution being published open access in 2020.

5The platform openaccess.nl11 proposed yet another methodology to monitor the adoption of open access among researchers affiliated to Dutch universities, and here again the dynamics are consistent with the other countries, with a 62% threshold already reached last year. In conclusion, it is interesting of course not to precisely compare the data because they do not proceed from the same methodology. We see that even with three different methodologies we observe the same dynamic in three different European countries.

6However, behind the global dynamics it is necessary to dive into the details and try to identify the logic of models that support this movement. Several routes were identified quite soon in the history of open access, starting with the difference between depositing articles in open repositories, the so‑called Green route to open access, and articles published directly on the publisher’s platform, the so‑called Gold route. The landscape became even more complex when the hybrid concept emerged referring to articles published open access in non‑open access journals. Then of course there is the Diamond model, referring to Open Access (OA) journals, free for both readers and authors, meaning no charging article processing charges or subscriptions.

7What are the respective dynamics of those models of open access? The traditional opposition between the so‑called Green and Gold routes theorised by Stevan Harnad 20 years ago is becoming less and less relevant in the current context. Following the multiplication of policies and institutional strategies, there is a growing number of articles made available open access both on the publisher’s platform and in an open repository. This evolution, partly but not only due to the role of PubMed Central12, renders the long‑time debate between the two routes to open access somewhat obsolete. It opens the possibility to think of both routes not as in competition but rather as complementary to each other. Each route brings its own way to open access publishing practices. In my opinion this type of evolution is a sign of the maturity of community regarding OA.

8However, on the other side, despite multiple disincentives from funders and library consortia that consider hybrid publication as perpetuating the status quo and introducing the risks of double dipping in open access funding, the model is still present in the landscape as evidenced by the hybrid OA dashboard proposed by Najko Jahn13. Hybrid publication is indeed a flexible model that accommodate well with the contradictory requirements between the mandates to publish open access and the need of researchers to publish in the existing recognised journals of their discipline. In fact, hybrid is not a model. It is certainly a temporary adaptation to a transition period and that is probably the reason for its permanence until now.

9At the same time, the multiplication of so‑called transformative agreements (Tas) gives a solid push to hybrid publication14. The ESAC Registry15 that records most of the TAs gives evidence of this growth. This is really an important shift in the landscape with the tremendous growth of articles freed by the TAs according to ESAC figures. However, a recent study that surveyed 197 TAs concluded that globally those agreements do not provide any real transformation in the business model of journals, nor did transformative agreements help to flip these journals to open access. That is mainly because this is not in fact the primary intention of the stakeholders signing the contracts.

10Which model will survive and wipe the others out from the market? From what precedes it is already obvious that each model brings its own type of success and addresses a specific need in a specific situation, even temporarily. That is not, however, the end of it of course because recent years have brought into focus other open access publishing models that have slipped under the radar so far. The first one is of course the Diamond model, which designates those publications that are free for both readers and authors. In other words, they do not charge authors for publishing their articles in open access. A recent study brought to light the important contribution of this model to the global development of open access worldwide.

11Here again the reality is more complex than what we think, and it would be a mistake to consider Diamond and Diamond journals as an exclusive product of Latin America and social sciences and humanities, where it is indeed extremely important. In fact, what a study reveals16 is that the Diamond model is prevalent in many disciplines and most of the regions of the world, including Europe. In other words, while most of the open access discussions were focusing on the glamorous model of the publishing landscape, the Gold APC model, the vast majority in fact of open access journals were frankly adopting another model, the impact and importance of which was wrongly being overlooked by most observers.

12The Diamond journals represent an overlooked and neglected part of the OA landscape because by nature they are not involved in the most inflammatory debates related to the stronghold of commercial publishers over scholarly communication. Most of the time Diamond publishing venues are owned by scholars through their universities, research organisations and scholarly societies. In fact, they represent the quiet adaptation to the digital age of a long tradition of scholarly publication. They constitute a remarkably persistent model of open access publication with a stable number of articles published throughout the years, and in that sense the comparison with the evolution of articles published OA in exchange for payment of an APC is striking.

Fig. 5: Shares of OA Diamond and APC-based open access models in DOAJ-listed journals

13The Open Access Diamond publishing model stands outside of the capitalistic economy of for‑profit publishing and is therefore insensitive to market incentives. In addition, the OA Diamond publishing model does more than simply address the needs of researchers. It represents an idea of autonomy and self‑rule of the scholarship community in the new digital context.

14One of the gems uncovered by the Diamond OA study was this continued innovation throughout the whole sector, with a diversity of initiatives across different disciplines, languages and countries, with scholarly communities working constantly on inventing and adapting their communication practices to their needs, taking full advantage of the possibilities the open web offers. In the Diamond study already mentioned a series of open questions were submitted to almost 1 200 open access journal editors about their motivations. Ethical and political values appeared prominently as key drivers for their professional and personal investment in open access. I would like to share with you one of the most compelling answers we received in this study: ‘Our open‑living ethos is not just about open access, open software, open review. We are committed to living, discussing, deciding and failing in public. We discuss and develop our project values and policies openly as our public message’.

15Beyond the necessary complementarity of the different routes to open access, new types of debates have emerged in recent years around the concept of diversity in the open access movement. Several voices in the community have called out to take a step back and try to reflect more on the why and not only on the how of open access. They highlight that through its inclusion in institutionalised open science policies the open access movement could fail its original purpose at the very time it seems to be succeeding.

16It is true that driven by innovation the adoption of open science led to a strong diversification of technologies, formats and workflows in academic publishing, at least in appearance. The number of new services proposing new ways of creating, assessing, disseminating, discovering and reviewing research results is staggering. 20 years ago, Stevan Harnad’s modest proposal17 to radically transform scholarly communication by sky writing research on the Internet had many offspring: preprints, mega‑journals, digital editions, overlay journals, next gen repositories, publishing platforms, data papers, living books, micropublications and of course 50 shades of open peer review. This fascinating world of innovation in scholarly communication has been wonderfully well charted by our colleagues Jeroen Bosman and Bianca Kramer from Utrecht University with their project 101 Innovations18.

17However, at the same time, we have understood that open science and its powerful push for diversification through innovation does not prevent the concentration of power in the academic publishing system. Somehow, like the cyberlibertarian utopia of the open web suddenly reverted to the dreadful control of every means of communication by the GAFAM, many articles have shown that a limited number of commercial companies are in a position to control the entire academic knowledge production workflow through strategic mergers and acquisitions in the open science context. This led to a number of position papers and initiatives, such as the call for action that Kathleen Shearer published last year19 providing a number of recommendations to foster bibliodiversity in scholarly communication.

18This leads us to consider academic publishing from a different perspective. Our colleague Leslie Chan recently gave a striking presentation at the FORCE2021 conference20 to encourage us to challenge structural racism and systematic biases in knowledge production. In his talk he showed how paywalls are just the most visible part of a system of power in knowledge production that prevents the full adoption of open science practices. In other words, he encourages us to look beyond the different routes to open access, beyond the business models that support it, beyond the publications themselves and to start questioning the knowledge infrastructure that lies beneath the surface, including peer review practices, editorial practices and also the way in which the standards, the tools and the technology, nurture a monoculture of thinking in academia and stifle the diversity of epistemes that are essential in a true open science.

19As Leslie rightly points out, infrastructures have a crucial role in ensuring diversity or, at the opposite end, providing the backbone for monopolistic strategies leading to epistemic exclusion in scholarly communication. In recent years the role of infrastructures at all levels of open science have become more and more apparent. Because open access publishing is primarily digital publishing, even before being open, the question of sustainability is of utmost importance, but the question of openness applies here as well to ensure that the infrastructure does not recreate enclosures behind the door. It was well put by Geoffrey Bilder, Jennifer Lin and Cameron Neylon, the authors of the Principles of Open Scholarly Infrastructure21: ‘Everything we have gained by opening content and data will be under threat if we allow the enclosure of scholarly infrastructures’.

20Where do we stand after 30 years of development of open access, 20 years after the B declarations, 10 years after the tipping point? An observer coming back after 30 years of absence from the field would probably be horrified by the level of complexity it has reached since then. The multiplication of economic models, the continuous creation of new formats, new tools, new platforms as a consequence of permanent innovation and the intricacies of digital services interacting with each other apparently transformed academic publishing into a maze where it is easy to get lost.

21To better grasp the situation and identify the challenges before us I would like to propose, finally, to take a step back again and consider academic publishing from a different perspective. First, I would like to refer to a book published by one of my colleagues at the French School of Advanced Studies in the Social Sciences (EHESS), Valérie Tesnière. Her book Au bureau de la revue22 tells the long history of academic publishing in France during the 19^th and 20^th Centuries. This long history made of many extraordinary and diverse stories perfectly illustrates the true nature of academic publishing in its diversity, with an incredibly high number of different formats, practices and products constantly renewed throughout the decades. It is an history of constant innovation and creativity in connection with the changing needs of the scholarly community and society itself. This history reminds us that giving access to knowledge through publications is a longstanding effort that started long before the open access movement itself. The relatively short history of open access should be placed in the continuity of the longer history of scholarly communication.

22Second, I would like to refer to another book23 published by another of my colleagues at EHESS, Christian Jacob. In his research Christian studies what he calls places of knowledge, the places of knowledge of these all these objects, tools, frameworks, instruments that support the creation and the circulation of knowledge in various cultures and ages. Christian considers a place of knowledge as intimately linked to its social context and to other places of knowledge. In his view, knowledge cannot be considered independently from the culture and the society it comes. Knowledge should always be understood as part of the social fabric where it finds its meaning. My proposal to you is to consider academic publishing as a network of places of knowledge, a public garden of knowledge with many bifurcating paths, crossroads and meeting places that should be understood as such.

23The peculiar situation we are currently collectively living in led many of us to question the meaning of our activities. What are we doing exactly? Why are we doing it? What is important? As a social activity, knowledge production is not exempt from these kinds of questions. Our colleagues running the journals surveyed in the context of the Diamond study answered, in my opinion, very simply to that question: ‘Free access to scientific information is an essential element for economic and social development. We have in mind our firm commitment to the democratisation of knowledge. We should not put a price tag on ideas and public knowledge’. Therefore, the challenge ahead is quite easy to understand. It is not to guess what the future of publishing will be. It is not to bet on the future winner of the race between OA models and new technologies. It is not to declare obsolete traditional forms of scholarly communication that are essential to valued scholarly communities.

24The challenge ahead in my opinion is rather to efficiently support all those places of knowledge and continue to open them all for the benefit of society. We do need smart policies and strategies that support STEM and SSH, the journals and the books and the proceedings as well and the different languages in which researchers communicate with their peers and with society. We need to have the traditional and native publishing venues supported on their journey towards open access. We need to support those who want to experiment beyond the traditional formats as well as those who continue working on longstanding editorial collections in a cumulative way. We need to support diversity and to preserve it from predation as well. In other words, we need to increase and improve coordination across disciplines, communities and countries. We need to consider ourselves as partners rather than competitors, support each other, learn from one another and consider ourselves as serving the same goal each of us in our own way.

25Finally, I think we need a coordinated open science ecosystem where the different ethical and sustainable academic publishing practices are supported equally. They are the building blocks of the knowledge infrastructure, and we are not there yet.

To go a step further

26“The economic model of Diamond publishing is supported on three legs:

• The volunteering coming from researchers themselves, who invest a lot of their time and energy in running their journals, because that is so important for the scientific community.
• The strong institutional support provided by personnel seconded to these editorial venues by the institutions and the research‑performing organisations.
• Funding, coming from grants particularly.

27There is therefore in fact an economic model. It has been mostly invisible so far, so one of the outputs of this study was to bring light to this economic model, and now what we need is to have better and more efficient institutional strategies and policies to support more efficiently and in a more sustainable way this kind of economic model.”

28“The time spent by researchers on managing a scientific journal is necessarily time lost for research in the field. Therefore, it is fundamental that this editorial work be recognized as research work.”

29“Researchers are undoubtedly at the heart of Diamond Open Access, as pilots, but they are not alone. They are accompanied and supported in their editorial activities by robust and diversified infrastructures, which are sometimes complicated to talk about because of their technical nature. It is necessary to better understand and valorise these infrastructures and the indispensable role they play in the dissemination of knowledge, to support them financially and to pay attention to their governance, so that they best meet the needs of researchers.”