Diamond Open Access in Scholarly Publishing

Report submitted by Jean-Claude Guédon for review by the OPERAS advisory committee, aimed at progressing on the issue of governance for the Diamond option.

I. Introduction.

Designing the governance of any organization largely amounts to creating the organizational machinery that will achieve its goals while preserving it from potential dangers and threats. In the case of Diamond Open Access (DOA) in general, and DOA focused on Social Sciences and Humanities in particular, DOA must be seen as emerging within both an existing ecosystem of scholarly and scientific publishing, as well as, more specifically, an ecosystem of Open Access (OA) publishing. The context can be extended further and made more complex: marshaling the concept of “infrastructure” will help analyze and interpret the positioning of DOA as well as its dynamic elements.

The very notion of DOA – i.e. a way to publish scholarly materials that abstains from collecting money from readers (or their proxies, such as libraries) or from researchers (either from authors or their proxies, such as research institutions or research funding agencies) – points to the need of thinking about its financial support in terms that do not reconcile easily with the notion of a “business plan”. Indeed, business plans refer to “business”, and the use of this term points to a perspective where business concerns must cohabit with scholarly concerns, like the production of validated knowledge. The absence of revenue streams coming from researchers in whatever role – reader or writer – severely constrains the possibilities of business plans, a fact that has been tacitly admitted by some think tanks close to the commercial publishing sector when they began renaming “DOA” as “sponsored publishing”. This can be read as code for the impossibility of designing a viable business plan for DOA (which is exactly the point). Similarly, DOA has been promoted in various quarters not only as a way to solve a number of inequities in the scholarly publishing world – access inequities and participation inequities both mapping onto wealth inequities as they appear at very social scales – but also as a way to limit commercial encroachments into the knowledge-producing worlds to controllable limits.

Designing the governance of DOA cannot be thought up in isolation to everything else. Effectively, DOA is a programme with objectives that seek to optimize the processes accompanying the production of knowledge, and this means setting up these processes with as little interference from other sectors of human activity as possible while also responding as well as possible to human requests for validated knowledge. Many other programmes exist that support the production, dissemination, preservation and evaluation of knowledge. Most of them involve some kind of business plan, be it a subscription scheme (S), an article processing charge (APC), or a mixture of both. Combining these terms has led to a large number of transactional protocols. That very variety can be the symptom of an incapacity to think about knowledge production outside of a business framework; apparently, for many protagonists in scholarly publishing, particularly among publishers, if something costs something – and scholarly publishing certainly does – the only model available to meet the costs is a business model. Otherwise, it has to be “sponsored” or, in more abrasive terms, it is dependent on some charity. Therefore, it is not “sustainable”.

The previous paragraph aims at showing that DOA and its governance need to be thought up in the context of a situation that is already well established, with a majority of powerful players already in place. It is consequently important to remember that the governance of DOA, while respecting the best values of inclusiveness, democratic organizations, respect for differences, etc., must also be designed to resist the onslaught of hostile forces and institutions, and it must hold the organizational means to map out the strategies needed to chart a path forward that, ultimately, will reach the scale of the planet.

If the governance of DOA is designed in this spirit, i.e. in the context of a dynamic and constantly evolving situation, it becomes obvious that looking at the forces and events that produced our present situation is necessary if we want to forge ahead in as lucid and as thoughtful a manner as possible. This is not a matter of prophecy, even though good knowledge is here to help us make educated guesses about the future: rather, it is a matter of strategy design. To come back to the initial formulation, what kind of organizational machinery is best suited to the design of the most promising strategies for DOA? In short, this effort bears some resemblance to the drafting of a constitution that seeks to optimize certain elements needed for the guidance of as good a society as possible. In this case, we want to draft the DOA constitution that will maximize its chances of survival while optimizing the objectives that DOA seeks to achieve for knowledge and its production.

One last point before embarking on this journey: this document is resolutely knowledge- and research-centric. A desire to destroy anything that is already present is not part of this document. However, when knowledge and its purposes (theoretical, practical, regional, situated in other ways) come into conflict with existing organizations or institutions that command other agendas, the choice is simple and clear: knowledge and researchers will be the defining and commanding terms. An example can quickly illustrate this issue: researchers, to produce knowledge, need to publish. To publish actually covers a number of functions – registration, certification, preservation, dissemination – which are often and summarily devolved to the “publisher” category resulting in, from that perspective, the essence of publishing falling into the hands of publishers, and not the hands of researcher communities. The present document will choose “publishing functions” over “publishers” and will use this distinction to offer ways of reorganizing the publishing functions in ways that agree more with other elements of the publishing ecosystem.

II. How scientific publishing has evolved since 1870 – a historical sketch

The history of scholarly publishing is characterized by a constant evolution that must never be forgotten if we do not want to make the classical mistakes of “Whig” history – i.e. projecting parameters from the present into the past. A science journal in the 17th century in Europe is very different from a scientific publication presently, even though the term “journal” was already in use in the 17th century. But, in the 17th century, it would have been difficult to distinguish between a so-called “primary” journal where original research results are published from a gazette specializing in scientific or scholarly news or from a publication trying to popularize esoteric knowledge into a more accessible form to reach a wider public. Periodicals in that century tended to address all these genres under one single cover. Specialization appeared only gradually and slowly[1].

Beside their functions, journals also differ by their institutional status : purely commercial journals co-exist with journals emanating and supported by academies, both national and regional. Since the late 18th century, associations, both public and private, had begun to respond to various communication needs of their members by creating journals that, in effect, acted as the “bullhorn” of the society. Such journals were also exchanged or bartered for equivalent publications from other similar associations elsewhere in the country, and increasingly beyond, thereby contributing to the weaving together of a communication system in scholarship which, in truth, had already begun to exist around problems and conflicts as early as the Renaissance; Anthony Grafton has identified such communities structured and maintained by conflicts, such as issues of chronology[2]. They did not wait for “scientific” concerns to emerge as communities, albeit conflicted or not.

The professionalization of science in the late nineteenth century and the growth of research communities in the same century led to a stricter alignment of journals with emerging disciplines such as chemistry. Societies increasingly distinguished themselves between “scientific” associations and “professional” associations. Gradually, the former tended to relegate professional matters to the background while the latter continued to offer a rich variety of publications that ranged from pure research to different issues relating to working conditions of various specialists, in particular engineers.

With the growth of scientific publications in all kinds of forms, finding one’s way into the literature became increasingly difficult. In response to this problem, at least two solutions emerged. On the one hand, books were written to offer an encyclopedic, review-like coverage of a subject area. These publications, written by researchers, were quickly identified and promoted by commercial publishers. Unlike research results that were accessible only to cutting-edge specialists, these publications quickly became the needed tools to explore or study a new field, or to make sure that no relevant insights coming from somewhere were missed. When such tools began to reach beyond national borders and beyond a single language, the “internationalism” of research results was being increasingly foregrounded[3]. Of course, the possibility of being included or not in a review, encyclopedia or bibliography had practical consequences on the visibility of the corresponding publications, but such feedback mechanisms were not immediately perceived, interpreted, and even less exploited to stimulate sales of particular journals. Some publications were not included out of “benign neglect”, not because of strategic intents. However, the possibility of including or excluding was present from the start and would only gain in importance as time went on. Eventually, it would converge with notions of “excellence” and “elitism”.

The existence of several languages in scholarly publishing also meant that the editors of a particular journal in one language could routinely identify a significant article in another language, and have that article translated to publish it anew in their own journal. Serious constraints on such a practice did not manifest itself much before 1900 in Europe, when international copyright agreements began to take effect. By that date, it can be said that the North Atlantic region had developed a system of scholarly communication and publishing that also included Japan and certainly reached into Eastern Europe and Russia.

In the same period of the late nineteenth century, a new phenomenon was beginning to emerge: the rise of commercial publishers. Initially, commercial publishers had found publishing scientific or even scholarly journals quite challenging. The reason is simple: research results publications are highly focused, often quite esoteric, and legible only by a limited number of individuals. This was just as true of physics as it was of the humanities or social sciences. However, professional publications were more interesting because of their extensive preoccupation with working conditions, association news, etc. And associations were not always very good at the practical management of publishing processes. It is in this manner that Springer, after the Franco-Prussian war, began to offer services to various professional societies in forestry, pharmacy and engineering. This first inroads into scholarly publishing allowed Springer to demonstrate a win-win strategy with an engineering society by offering their expertise in print publishing as a service: journals were published on time, laid out more professionally, and the marketing of such publications was done far more extensively and efficiently than had been the case with a professional association[4]. Other publishing houses in Germany followed suit, thereby contributing to the gradual creation of a new scholarly publishing ecosystem.

Commercial publishers, eager to find new areas to exploit their technical and commercial knowhow, additionally began to explore new fields. As societies tended to be a little slow in following the emergence of the newest technologies such as electricity and radio, these publishers began to explore the possibilities of creating journals in such areas and make a profit from them. In the case of electricity and electrotechnics, Springer became sufficiently familiar with the publishing initiatives that had begun in the USA to desire imitating some of them in Germany. He therefore initiated himself in the international commerce of scholarly materials – an activity that would become extremely profitable later on, between the two World Wars.

By the beginning of the First World War, the German scientific publishing industry had become the most vital and the most visible in the world. Several languages played an essential role in the developing communication system of science, but German was the most important. German publications were also the most visible.

Being conscious of the strategic importance of scientific research certainly did not begin with the First World War, but the conflict brought that degree of consciousness to new heights. The allied nations then tried to rein in the German advantages in scientific research by attempting to fence it: international congresses, for example, were closed to German scientists. By avoiding to submit papers to German publications, some researchers from the “Allied nations” hoped to decrease their prominence[5]. This strategy proved ineffective, and its failure indirectly helped to demonstrate anew the importance of German research and scholarly publishing. By the late 1920’s, Germany’s scientific publications had regained their dominant status.

The destruction of the German scholarly publishing industry was largely the result of the racist policies of the Nazi regime: they either led or forced some of the main leaders of that sector to flee out of the country. These leaders took refuge in Holland, Britain and the USA. The war itself completed the destruction of the publishing industry. By 1945, the opportunity was open to transfer the German dominant position to other parts of the world. It is in this context that Robert Maxwell initiated himself to the mysteries of scholarly publishing and its international commerce. In a context where all the victorious nations were engaged in intense intelligence gathering to ensure the control of the best German technical innovations, transferring the publishing and commercial know-how of such a firm as Springer was obviously necessary.

Bartering some of the resources needed for Springer to operate against the knowhow and stocks of publications that had failed to reach the Western nations during the war, Maxwell started to follow the tactics that the German scientific publishers had developed so successfully in the late 19th century and the first half of the 20th. In this he was deeply helped – as were others, such as Elsevier – by the extremely rapid expansion of scientific research, now seen as a key strategic element in the context of the Cold War. Like Springer in the late 19th century, Maxwell with Pergamon Press, Elsevier, and a few other firms began to explore the commercial possibilities of journals that scientific associations or societies did not want to deal with: interdisciplinary domains such as bio-chemistry, new technologies, etc[6]. They also began to recruit scientists with strong reputations but which, for one reason or another, had never been able to penetrate the closed societies of learned societies and their strict hierarchies. To accomplish this, Maxwell was mentored by an old Springer hand, Paul Rosbaud[7] who essentially taught him all that had to be known to run a successful business in a complicated world. In particular, Rosbaud taught him to be attentive to the elements entering into the constitution of a publication reputation: prominence of the editorial board and the editor-in-chief were key points in this regard. But this was still only a “reputation”; the idea of rankings is not yet present or even implicit. Besides, in a world of fast-growing research everywhere, with expanding library budgets, a good reputation was enough to ensure the commercial success of a “primary journal”.

The explosion of the Soviet H-bomb shortly after the American test of a similar weapon, in 1953, and the launch of Sputnik four years later generated a kind of panic in the USA. One of the main themes underpinning the analyses that followed the panic was that US research suffered from a communication problem. One of the solutions supported by a committee organized to address this communicational problem[8] was an idea expressed in Science in 1955 by Eugene Garfield: a citation index[9].

The Garfield citation idea of 1955 is quite different from the Garfield perspectives on the citation index in the 1970s. In his first “role”, Garfield was looking for a better way to follow scientific information across disciplinary boundaries, pushed ahead in this direction by Joshua Lederberg, a Nobel Prize winner at the ripe age of 33. Interdisciplinary research was poorly covered by the existing bibliographies, indices and reviews. However, Garfield also had the contours of a businessman, and this meant that he had to find ways to sell his citation index.

Garfield’s first commercial success was Current Contents, a publication he launched in 1955 and which harkened back to the German practice of publishing reviews, abstracts and other peripheral publications to help researchers find their way in a vast corpus of literature. Current Contents is what allowed Garfield to finance other types of publications, including the Science Citation Index which began in earnest in 1964. It also allowed him to notice what German publishers had already noticed before the war – namely that the inclusion of a journal in his Current Contents had beneficial consequences for the subscribing readership of the journal. Libraries were paying attention to those lists in building their periodical collections, particularly in the USA.

Turning to social scientists, such as the sociologist Robert K. Merton and the physicist turned historian of science Derek De Solla Price, Garfield began to understand that his science citation index also held some potential to “evaluate” research results themselves. To cut to the chase, Garfield after 1972, began to see that his newly developed impact factor could help rank journals. In a crowning moment, Price managed to write a pithy text where rankings and the overwhelming importance of a minimal core of titles came to bring about a system of elitism, soon to be named “excellence”, that could also inform the concerns of any journal editor and his/her publishers[10]. This meant that the impact factor could become a fundamental tool both to do research on scientific communication, but also to structure the rules of commercial competition among journals. Garfield’s publication, Journal Citation Reports (JCR) began in 1975 and, gradually, its import became obvious: the whole market of journals reorganized itself, step by step, around this metric while the behaviour of researchers, in parallel, also began to conform to the constraints of the Garfield metric[11].

This remarkable convergence of commercial and intellectual concerns is sufficiently important to stop a minute and reflect about it. The best way to do so is to make a detour through the concept of “infrastructure” initially developed by Susan L. Star and her colleagues[12]. The reason is the following: the historical sketch that precedes can be interpreted as the step-by-step evolution of an infrastructure. The terms within the infrastructure tend to remain stable, thereby hiding the changes in their meaning, but journals competing within a quantitative framework – the rankings based on the impact factor – that purports to provide some insight into their “quality” or “value” are not the same as the journals being offered one library after another on the basis of their “reputation”.

III. What is a scientific publishing infrastructure?

1. Infrastructure

Elena Šimukovič’s doctoral thesis is essential on this point. She reminds us that the defining features of an infrastructure are:

  1. Embeddedness within other structures, social arrangements, and technologies;

  2. Use transparency (or invisibility) always already “at hand”;

  3. A reach that goes beyond single events or single sites;

  4. Linked with a community of practice which is shaped by the infrastructure while also shaping it (e.g. the legacy of typewriter keyboards in the design of computer keyboards);

  5. Must be learned somehow as the condition for being a member of a community of practice;

  6. Embodiment of standards (especially when infrastructure “plugs into” another infrastructure).

  7. Built on an installed base which also acts as a form of inertia that the infrastructure. This tension requires a concern for “backward compatibility”;

  8. Visible only when it (begins to) breaks down, including its politics and its organization;

  9. Its fixity or stability is tied to modular increments that are neither global or occurring all at once. This layered and complex structure implies that it cannot be changed from above, but only through adjustments with other elements of “involved systems”, and through negotiations. In the last analysis, nobody is really in charge of an infrastructure.

  10. It is probably better to approach infrastructure as a verb rather than as a noun[13]

2. A scholarly publishing infrastructure

The infrastructural approach helps us reinterpret the historical sketch that has just been quickly exposed in the previous pages, and look at contemporary events in an interesting light. First the reinterpretation:

  1. Scholarly publishing is part of publishing, yet distinct. This is reflected in the fact that directly solving the commercialization of scientific journals proved difficult. Steps in that direction included a gradual shift of emphasis from book publishing to periodical publishing (e.g. Springer in the first half of the 20th century), as well as building categories of “meta-publications” such as indices, bibliographies, abstracts, reviews, etc. It also involved working first with domains at the boundary of scientific publishing – professional and engineering societies – with publications dealing with a lot more than pure research (professional life, economic and legal issues, etc.). In the first years of this transition (end of the 19th century in the case of Springer), initial steps were taken in areas of the local publishing industry that were least developed (specialized publishing about new technologies such as electricity emerged more quickly in North America, for example). The contact with the USA, in the case of Springer, stimulates the growth of international commerce in publishing along with the know-how associated with it.

  2. Publishing in periodicals is already common and familiar to researchers after 1870. Relying on local societies is the most familiar route to publication. Publishing in society journals appears as little more than an extension of research and professional relationships.

  3. Publishing, as a generic function, extends beyond one event and one institution. Every researcher eventually comes to the conclusion that research results do not really exist if they are not published.

  4. Scientific periodicals evolved not by radical breaks with the past; they evolved by intensifying the specialization of their functions (primary journals, disciplinary base, scientific and technical news, popularization of research, etc.). Society journals belonged to communities of practice. An innovation tested in one society tended to diffuse gradually among other societies and their journals.

  5. Scholarly and scientific periodicals plug into distribution circuits (bookstores, etc.) and preservation systems (libraries) that already exist. They fit within the standards, habits, laws that govern their activities.

  6. Scholarly and scientific journals emerge and grow in the midst of already existing journals which act as an installed base with its specific technical, financial and legal frameworks.

  7. Researchers look for information. Journals may help them find that information, but the information may be pursued across several journals. Various tools help researchers find how to obtain the desired information. Journals not identified as journals directly subscribed to by oneself or one’s institution “fail” if they are hard to identify, locate and access. This experience is lived as a failure that affects fundamental processes. As a result the situation must be corrected, which means improving the infrastructure. In the same movement, the level of infrastructure has become visible, which is what happens when it fails. The creation of the International Catalogue of Scientific Literature after 1900 illustrates an attempt to correct such an infrastructural shortcoming.

  8. The historical sketch provided above illustrates how changes occur by bits, in a modular fashion, as well as their local nature. The commercialization of journals did not start with scientific societies; instead, it started with technical and professional societies. Then it gradually spread to reach scientific societies. Internationalizing the commerce of scientific journals came as an unanticipated consequence of the desire to take position in areas of publishing neglected locally in Germany, etc. The same is true after WWII. Some firms (Elsevier, Pergamon Press) explore areas neglected by the societies, in particular interdisciplinary domains (biochemistry for example, or genetics) or new technologies (atomic energy, radar, etc.). Likewise, the evaluation elements of the existing scholarly and scientific publishing elements gradually shifted from a fuzzy, reputational, economy to a quantified ranking system based on citation indexing. However, the ranking device stemmed from an attempt at mining the existing literature at the level of ideas, a level linked to citations. The process was not conducted lucidly or with foresight. In trying to make his Science Citation Index financially viable, Garfield evolved its functions. The people he consulted both helped and warned him about possible downsides (e.g. R. K Merton and the risks of using citation counts to evaluate people), while others provided formulations that intensified the competitive and elitist dimensions of the incipient ranking system (D. DeSolla Price). Humans can affect infrastructures through their intentions, but it is much more difficult for them to predict the consequences of their actions. These may diverge very significantly from their original intentions.

  9. All that precedes points to an on-going, never-stopping, process at work within an infrastructure. Verbs are better suited to this task than nouns. We are still “infrastructuring” scholarly publishing, but we need to identify the pressure points of the system if we want to affect its evolution in some fashion.

Infrastructures are less things than they are processes “... where socio-technical relations are formed and maintained.”[14] and they involve complex networks of actors, both human and non-human, congruent with the Actor Network Theory[15]. Infrastructures can also be approached as a kind of socio-technical “baggage” endowed with a variety of affordances, some of which open future possibilities, while others tend to act as an inertial drag that maintains elements of the past. However, they also show that envisioning the governance of infrastructures is difficult: the inherent complexity of the concept stands in the way. This is fair warning for anyone trying to conceive or design the governance of something as infrastructural as “Diamond scholarly publishing”.

3. The technological shift: the changing status of journals

The presence of journals has remained constant during the whole period (ca. 1870 – present) sketched above, even though their functions, natures and roles have constantly changed. This is important, because, at present, journals are more present than ever, and we can anticipate that the terms are going to be used for a fairly long time. However, they will change again and the analysis presented so far has neglected the drastic technological shift brought about by computers networked on a planetary scale.

While the role of journals has been enormously increased by an evaluation  system of research that relies on their ranking through impact factors, this evolution has also given rise to a few paradoxes. In particular, the further convergence of rankings with elitism or excellence has led to a problem for commercial publishers: if they argue too strenuously in favour of competitive excellence, and if they succeed, libraries may well take Derek DeSolla Price’s argument at face value and decide that one can obtain half of the “value” – whatever that is – of a set journal titles by buying the square root of the number of titles[16]. The “long tail” that forms most of the journals in any publisher collection is thus devalued to a significant extent.

In response to this problem, commercial publishers have relied on two important elements. For one, they have benefited from the time delay needed to implement the new, evolving, characteristics of the publishing infrastructure on a large scale. As the science fiction writer, William Gibson, famously stated: “The future is already here – it’s just not very evenly distributed”. The impact factor was gaining in acceptability in the ’80’s, but old habits die hard and libraries, despite growing budgetary problems in that period, simply could not drastically cut the number of subscriptions by following Price’s analysis to the letter. Bradford’s law remained dominant among librarians[17]. But the threat kept nagging publishers: libraries might severely curtail the long tail of their subscriptions, and some mitigating measure was needed.

The transition to digitization and the Internet provided the second element needed to protect commercial publishers. Journals could be digitized and set up on what were first called “portals”. Managing a portal was the new way to manage subscriptions, and immediately, a new idea emerged: since libraries complained about the prices of journals rising too much too quickly, publishers could open the whole collection at a flat rate, with the following results:

  1. All the journals are sold regardless of where they rank in Garfield’s JCR. This is the paradox: publishers rely on the rankings of JCR to set the price of individual titles, or the price of APCs for articles of a specific journal, but they sell everything as if these titles had equal value. This is known as the “Big Deal”;

  2. A discount is applied for the cost of the whole collection of the “Big Deal” as compared to the sum of individual subscription prices; yet publishers can still increase their revenue stream because they sell more titles, and they muddy the water by pointing out that the price per title has gone down. Of course, what imports for the library is the cost per actual use, but that part of the discussion was somehow swept under the rug and it has remained in the background for a long time[18].

  3. The increased revenue stream for a large collection of journals means that a greater share of the library budget goes to the larger publishers. Smaller publishers tend to be squeezed out, and this trend can increase the concentration of the industry.

  4. The presence of a portal is also useful for the “long tail” of journals: anyone seeking a particular article can also be steered in the direction of other, related, articles present on the portal. In this manner, the attention of the reader is retained longer, and this increases the probability of further requests to the portal with positive consequences for the statistics of the site and, indirectly, for the IFs of the journals in the long tail. This latter point, incidentally, directly relates to the “economy of attention” so central to Internet-based transactions.

All these details matter, to be sure, but they also point to a very important conclusion: journals, while evolving, have been erected into fundamental underpinnings for both the evaluation of the intellectual value of research results and the commercial value of the objects being traded. Caught in the ranking system that the impact factor has made possible, journals nowadays are punching well over their actual weight. This vulnerability is indirectly demonstrated by strategies that contradict the present status of journals: the “Big Deals” and their equivalent expression in an OA context; the methods used to retain the attention of a researcher on a particular portal or platform; the very existence of the citation index that points researchers to specific articles, not journals. Journals are now deeply embedded within the infrastructure of scientific communication and publication and, step by step, they continue to evolve within their new ecosystem – an ecosystem dominated by the impact factor and by the oligopolistic structure of the commercial firms. An example of the recent journal evolution is the “cascading” system used within a single stable of titles owned by a particular company: a submitted article is peer-reviewed and, if not accepted at the top-ranked publication, it is directed to a lower-ranked publication with no further peer review, only desk editors, and so on until it finds a suitable level for publication. In such a scheme, a journal belongs to a hierarchical family of titles that work in concert within the general commercial logic of the owning firm. The article is placed at the journal level where it fulfills its commercial function best.

Such examples demonstrating the constant evolution of the infrastructure are useful to provide some ideas about ways to act upon an infrastructure. Commercial publishers have succeeded in taking a commanding position little by little, sometimes in a sleep-walking manner, and over many, many years – decades in fact – by constantly reworking the essence and the nature of the journal to make it fit their tactical needs better. The constant goal has been to transform the journal into a better merchandise by gradually emphasizing its role as an instrument of evaluation rather than one of communication, discussion and debate. DOA should work at the same level to recover the lost ground and extend it thanks to the new affordances of the networked digital sphere. DOA should, therefore, focus on journals in their relationships to platforms to align them better with research in general, and the DOA project in particular. In order to achieve this result, it is important to remember that, in a digital context, the very existence and the meaning of journals stems directly from the platform, including its formats and algorithms.

The first (and most urgent) task of DOA is to decouple journals from evaluation purposes. Evaluation of research rests on a wide range of criteria which differ with each evaluator. Making it rest on a single and particular form of journal behaviour – namely, its ability to garner citations from a selected club of other journals – is both flawed and simplistic, not to say completely stupid. A single quality indicator, however, managed by the right kind of institution – a commercial outfit such as, nowadays, CLARIVATE – has become a powerful tool to shape market rules that are under the control of a few companies depending on the existence of a single (complicit) evaluator. It has led to a “globalized” transactional system that clearly advantages certain industrial groups at the expense of research itself.

If journals are clearly separated from the impact factor, they can exist only if they relate directly to groups of researchers that are coming together thanks to a variety of means: a common problem, a common instrument, a shared controversy, an urgent situation to address (e.g. COVID), a common problem framework leading to a family of questions (deconstruction, feminism, decolonization, etc.). They can also be the instrument an institution uses to project its voice, as The Philosophical Transactions of the Royal Society demonstrates. In other words, many journal models exist that stem from groups of researchers, not from publishers. Incidentally, neither Nature nor PLOS One fit this description and, in fact, they exemplify publications that begin to act like a platform.

DOA needs to work on models of platforms that can help navigate the vast sea of research materials available first on the platform itself, and then, of course, all the other collaborating networked platforms. Journals can be helpful in this regard because they provide a sort of rough categorical cut to define categories of knowledge and research groups. At thew same time, this formulation points to the primitive nature of such classifications: in contrast with the possibilities of mining and clustering texts, organizing knowledge through journal titles appears crude at best.

4. What platforms can do

Nowadays, the usage of the term “portal” has regressed in favour of “platforms”. Alas, platforms such as Facebook (now META) and Twitter (now X) have generated many legitimate worries: the toxic environment of fake news, echo chambers, tribal behaviour, trolling, harassment and intimidation is on everybody’s mind when the word “platform” appears[19]. However, the fundamental problem lies more with “platform capitalism” than with the concept of platform per se. Platform capitalism leads to “surveillance capitalism” and, as a result, platforms have been received with a certain degree of suspicion, as instruments designed to monitor and control researchers[20]. Platform capitalism is just a sub-genre of platforms in general, and platforms different from Meta and X exist. The danger, as will be seen later, is real, but, as with all technological objects, it is not because they can be wielded dangerously that they are dangerous. This is a perspective that also calls for some remarks at the level of governance.

How users of scholarly materials relate to platforms in both producing and making use of scholarly materials, and how these materials relate to each other through the various functions that a platform provides (citations, links, summaries, translations, text mining, texts and source documents, etc.) are parameters that deeply characterize a platform. How documents are evaluated, valued, diffused further, can also be influenced by platforms, hence the critical importance of their design. At present, for example, numerous platforms of commercial publishers mix a variety of criteria ranging from the legitimate functions of research materials to the promotion of certain journal titles, research themes, research sites and other parameters that serve financial and perhaps other objectives as well. The latter objectives reach beyond the research environments, and may even distort and warp them.

In the scholarly world, platforms should ensure the credibility, reliability and quality of what they hold while providing tools to researchers (and tools to build new tools) that allow the best use of the published units, be they notes, data, software, comments, articles and even monographs: easy document mining, analyzing, organizing in meaningful clusters, comparing, translating, etc. are some of the tasks platforms can take on locally and at large.

The pernicious power of platforms such as Meta or X comes from algorithms that stimulate user engagement and create human tribes rather than networks. Scholarly platforms obviously conserve the “social network” ability of all platforms, but that ability should be designed to achieve the objectives of research communities, and not of profit-seeking companies. In other words, research platforms should also be designed to enhance and protect scholarly communication.

Electronic publishing depends on platforms and not journals; journals are ways of organizing editorial tasks, and in the print world, they strongly shape and inform editorial tasks. In the digital world, this is no longer the case as editorial tasks can be designed without journals and on the basis of publishable units conceived very differently[21]. Unlike portals that are little more than virtual shelves, platforms have the ability of deeply sculpting the way journals (and other kinds of documents such as data, software, monographs, etc.) may exist and interact with users ; journals, on the other hand, display no symmetrical ability to shape platforms. Platforms, therefore, are more fundamental than journals, and obviously form the foundation of electronic publishing and its governance.

This argument is as valid for DOA as it is for the commercial realm – and commercial firms have long known it. Their participation in various industry-wide standards such as “Digital Object Identifiers” (DOI)[22] indirectly reflects their understanding of this particular facet of the electronic publishing industry. The gradual emergence of the platform as the dominant element of electronic publishing, scholarly or not, underpins the paradox haunting commercial scholarly journals: after having been erected as the keystone element of the redesigned market structure, journals find themselves sitting within a platform where they are bound to play a secondary role in the heuristic strategies usually deployed by researchers. In other words, the convergence of intellectual and commercial value that rankings have been striving to achieve is already unraveling because the presence of platforms has made the object “journal” partially obsolete. How they are calculating to move beyond journals – surveillance capitalism strategies, anyone? - is certainly a question of some importance for the near future, and it is beginning to attract some attention[23].

As any infrastructure, platforms also have to fit within the wider Internet infrastructure. As will be seen, the problems that a DOA infrastructure and governance system must overcome bear some resemblance with the problems met by the Internet in its early years, except that the early Internet had the support of the US military while DOA does not benefit from any ally with such resources. On the other hand, the world’s expenditures on scholarly publishing is also much smaller than growing and maintaining vast, well-equipped, armies.

5. Lessons from the Internet

When computer scientists began to master the intricacies of computer networking, they did so in a disorganized fashion. Various projects in the United States (Leonard Kleinrock and ARPANET), in the United Kingdom (Donald W. Davies) and in France (Louis Pouzin) were clearly pointing to the need of thinking about inter-networks, but each project pushed on as if it were the only worthy one. DOA offers a similar landscape nowadays.

With the concept of an Internet, Cerf and Kahn moved beyond this parochial level by enlisting willing colleagues to identify constitutive elements common to any network. They also focused on how exchange procedures between networks could be conceived and take place, and how an address space could be evolved across an inter-network bound to evolve greatly and rapidly. The last element is absolutely crucial in a device with undefined boundaries.

The famous 1974 paper by Vinton Cerf and Robert Kahn[24] merits a pause to retrieve some of the concerns they were addressing:

By analogy, one can say that:

The early history of the Internet also suggests a number of first steps that are obviously crucial for the governance issues.

  1. The important, sine qua non, actors of the Internet were the emerging networks. The important actors of DOA are the platforms, not journals. These platforms may not even deal with journals at all (e.g ArXiv, or similar platforms dealing, in the case of OPERAS, with SSH). Any attempt at governance of DOA must start with some kind of constitutional assembly of an initial number of platforms.

  2. Technical issues relating to platforms must be treated separately. Some are purely internal to a platform and should not become the concern of other platforms. For example, the technical side of the editorial treatment of submitted “publishable units” – in practice articles – by the OJS collection of software need concern only a particular platform. It may not even be found in another OJS platform because the OJS suite of software is configured in a particular way on each platform.

  1. There must be fulfilment of the basic publishing functions: registration, certification, preservation and dissemination.
  1. Enough has been said here to adumbrate the tasks at hand, and how to approach them. The point here is not to offer solutions to every possible problem, but rather to offer approaches that have a chance of moving the whole infrastructure in the right direction while protecting it from outside threats.

6. What should scholarly platforms prioritize?

Again, the priorities addressed here should be read as illustrations of the tasks at hand rather than a full blueprint of the ultimate organization of platforms – a stage which may not be reached for another forty or fifty years in any case, if we rely on the time scale needed to transform the infrastructure of scientific publishing from a society-dominated system to a competitive market structured by the impact factor. That being said, the priorities targeted by the envisioned network of platforms will focus on many of the failings of the present Open Science movement, failings largely due to a lack of comprehensive understanding of what scientific communication and publication is, and how power is working in its midst. In other words, the Open Science movement has behaved a little too naively and innocently to find the strategies needed to retake the initiative and defend itself against the encroachments of commercial entities.

(Etc. etc. As stated earlier, this is not a blueprint document, but rather the feeble attempt to begin articulating a perspective on DOA.)

III. Conclusion

The governance of DOA must be thought out as a distributed structure and a series of guiding principles that will help to achieve its objectives – namely a system of scholarly publishing and communication that will be under the control of research actors – researchers, research managers, research funders – and that will be completely embedded in the research process.

Research costs money, a lot of money, but the economic, strategic and social reasons for carrying out research are perfectly clear. That is why researchers have been supported at various levels of intensity, probably as early as the alchemists who were trying to produce gold for their king – a practice still alive at the time of Kepler.  As a result, the money side is debated as to its magnitude and distribution targets, not its existence. It is not a question of “business plans”, but rather of financing schemes[31]. If scholarly communicating and publishing are fully embedded within the research process – “Research is not complete until it is published” was the position of the NSF shortly after its foundation – publishing financing will be submitted to the same challenges as research publishing in general, but the source of its financing will not be questioned[32]. Therefore, instead of being stuck within a “business model” quandary, these financial schemes will become part of a policy.

Platforms are the dominant objects of the digital publishing landscape. While covering audiences that are not limited to countries, they can and do relate with the political structure of nation states. A network of platforms clearly reaches beyond any country or region to reach a planetary scale. This is what the network of DOA platforms should aim for. At the same time, its supra-national ambition – an ambition that is inherent rather than expressed in such a project – means that it must also position itself next to, or within existing supra-national bodies. In this regard, the early collaboration of DOA and UNESCO is a promising element so long as it does not evolve into an operation aiming at ensuring some form of control over the process. The final objective of a DOA governance is to nurture, encourage, welcome and help all those genuinely interested in such a project while providing enough structure, discussion spaces and implementation possibilities to help DOA become the default mode of scholarly publishing within the next fifty years.

[1] Csiszar, Alex. The Scientific Journal: Authorship and the Politics of Knowledge in the Nineteenth Century. University of Chicago Press, 2018. Tesnière, Valérie. Au Bureau de la Revue: Une Histoire de la Publication Scientifique (XIXe-XXe Siècle). Éditions EHESS, 2021. pp. 29-32.
[2] Grafton, Anthony. Worlds Made by Words: Scholarship and Community in the Modern West. Harvard Univ. Press/Triliteral, 2009.
[3] A fuller development would need to include the impact of the international exhibitions, starting with London in 1851, and the development of international conferences in the margins of these spectacular events.
[4] Sarkowski, Heinz. Springer-Verlag: History of a Scientific Publishing House. Springer, 1996, vol. 1.
[5] Brigitte Schroeder-Gudehus. Les Scientifiques et la Paix: La Communauté Scientifique Internationale au Cours des Années 20. Presses de l’Université de Montréal, 1978.
[6] Robert N. Miranda, “Robert Maxwell : Forty-four Years as a Publisher”, in Fredriksson, Einar H. A, ed., Century of Science Publishing: A Collection of Essays. IOS Press, 2001, pp. 77-89.
[7] Arnold Kramish, The Griffin. Macmillan, 1987
[8] Alvin N. Weinberg & alii, « Science, Government, and Information : The Responsibilities of the Technical Community and the Government in the Transfer of Information », President’s Science Advisory Committee, janvier 1963.
[9] Eugene Garfield, « Citation Indexes for Science. A New Dimension in Documentation Through Association of Ideas », Science, 122 (3159), juillet 1955, pp. 108-11.
[10] Derek J. De Solla Price, « Some Remarks on Elitism in Information and the Invisible College Phenomenon in Science », Journal of the American Society for Information Science, 22 (2), 1971, pp. 74-5, 74.
[11] Paul Wouters, « The Citation Culture », Doctoral Thesis, University of Amsterdam, 1999. 1.
[12] For our purpose here, the best introduction we can use is our colleague’s doctoral thesis, Elena Šimukovič, « Of Hopes, Villains, and Trojan Horses: Open Access, Academic Publishing and its Battlefields” University of Vienna (Austria), 2023. https://digitalcollection.zhaw.ch/bitstream/11475/28350/3/2023_Simukovic-PhD-thesis-OA-battlefields.pdf (verified October 3rd, 2024).
[13] Grisot, M., & Vassilakopoulou, P. (2017). “Re-Infrastructuring for eHealth: Dealing with Turns in Infrastructure Development.” Computer Supported Cooperative Work (CSCW), 26(1-2), 7-31. https://doi.org/10.1007/s10606-017-9264-2 . Cited by Elena Šimukovič.
[14] Šimukovič, op. cit., p. 43, quoting M. Grisot and P. Vassilakopoulou.
[15] The English version of Wikipedia provides a useful introduction of the Actor Network Theory or ANT.
[16] See note 9.
[17] If N articles of value (for someone or a specific group) are found in X journals, doubling N will require consulting X 2 journals. See https://en.wikipedia.org/wiki/Bradford’s_law.\ [18] The librarian of MIT, Chris Bourg has cancelled all subscriptions to Elsevier journals while promising the members of her institution that she would pay for every individual request and would provide the needed documents in a very short time. See https://libraries.mit.edu/scholarly/publishing/how-to-access-elsevier-articles/. Significant savings have been achieved.
[19] Frank A. Pasquale, “Two Narratives of Platform Capitalism”, Yale Law & Policy Review (2016). Available on-line at https://digitalcommons.law.umaryland.edu/fac_pubs/1582 . On p. 311, two opposite narratives of platform capitalism are clearly presented in tabular form.
[20] There is evidence to support this thesis. See “Landscape Analysis. The Changing Academic Publishing Industry – Implications for Academic Institutions”, SPARC report, 2019. https://infrastructure.sparcopen.org/landscape-analysis . More theoretically, Marion Fourcade has elaborated the related concept of “Ordinal Citizenship” published in The British Journal of Sociology (2021). https://doi.org/10.1111/1468-4446.12839 . The article is available at https://pure.mpg.de/rest/items/item_3318843/component/file_3375762/content\ [21] For some embryonic musings about this point, see Niels Stern, Jean-Claude Guédon, and Thomas Wiben Jensen. "Crystals of Knowledge Production. An Intercontinental Conversation about Open Science and the Humanities." Nordic Perspectives on Open Science 1 (2015): 1-24. https://septentrio.uit.no/index.php/nopos/article/download/3619/3454/. In a similar manner, the F1000Research platform rests on a notion of article that closely mimics the identification of software versions. https://f1000research.com/\ [22] Jonathan Clark, the present “Managing Agent” for the “DOI Foundation” has worked at Elsevier for 20 years. https://www.doi.org/the-foundation/board-and-governance/\ [23] SPARC, in the United States, has issued reports about Elsevier and Springer regarding surveillance issues. See “Navigating Risk in Vendor Data Privacy Practices: An Analysis of Elsevier’s ScienceDirect. » https://zenodo.org/records/10078610. “Navigating Risk in Vendor Data Privacy Practices: An Analysis of Springer Nature’s SpringerLink ». https://zenodo.org/records/13886473\ [24] V. Cerf, R. Kahn, “A Protocol for Packet Network Intercommunication”, IEEE Transactions on Communications, Vo, COM- 22, No. 5 (May 1974), 637-648. https://doi.org/10.1109/TCOM.1974.1092259 . A delightful statement needs to be introduced here as it applies so well to OA Diamond platforms: “Even though many different and complex problems must be solved in the design of an individual packet switching network, these problems are manifestly compounded when dissimilar networks are interconnected.”
[25] See https://en.wikipedia.org/wiki/Digital_preservation/\ [26] See https://en.wikipedia.org/wiki/LOCKSS/\ [27] https://en.wikipedia.org/wiki/CORE\_(research_service)/\ [28] “The p-value is the probability that your results would look the way they look, or would seem to show an even bigger effect, if the effect you’re interested in weren’t actually present », Stuart Ritchie, Science Fictions: How Fraud, Bias, Negligence, and Hype Undermine the Search for Truth. Henry Holt and Company, 2020, p. 87. The whole chapter on bias is highly receommended.
[29] https://openalex.org/\ [30] https://coar-repositories.org/what-we-do/notify/\ [31] For a particularly amusing example of business plan obsession, see Tasha Mellins-Cohen, “Classifying Open Access Business Models”, Insights, 37, 2024. https://digitalcommons.unl.edu/cgi/viewcontent.cgi?article=1334&context=scholcom. The DOA approach is baptized “sponsored models”, a term apparently invented by Delta Think. For a deeper understanding of what Delta Think is, see K. Strauch and T. Gilson, “ATG Interviews Ann Michael, President and Founder, Delta Think” Against the Grain, 29 (5), 2017, 14, pp. 40, 42-4. https://docs.lib.purdue.edu/cgi/viewcontent.cgi?article=7842&context=atg\ [32] J. Merton England, A Patron for Pure Science. The National Science Foundation’s Formative Years, 1954-7, NSF 1982, p. 175.

Creative Commons © 2024 Jean-Claude Guédon. This work is licensed under CC-BY 4.0.