Note: This is a preprint based on the same-titled speech in Fiesole Retreat 2024, Cape Town, South Africa, May 7-8, 2024, Session 2: Equity and Inclusion in the Journal Ecosystem: Access, Discovery and Exploration. It should not be considered as finished work.
Printable version at https://hdl.handle.net/10889/27035
Introductory notes
Paradoxography is recording and speaking about the paradoxes, the things that are marvelous and extraordinary. Quite often it is associated with the abnormal, the objects that are out of the order of the physical world and its stereotypical attributes, as well as those that challenge our senses and common reason. As a research field, it is met in the classical literature, mainly the Hellenistic one, and for the purpose of this speech the term symbolizes both the odds in the representation of the Greek language in the scholarly communication and the effects of the English language hegemony on this field of the humanities, as it will be later seen.
The current study is stimulated by discussions over the last two years regarding the bibliometric effect of certain disciplines in ranking Greek Universities. There has been a public debate in the country about the role of these systems in affecting the decisions for university funding. As the debate cascaded, there were certain questions about the visibility and recognition of the work of scholars from disciplines that opt for publishing in Greek and other languages or in formats that are not on the radar of the indexing databases.
Therefore, in this presentation we speak about the abnormalities (paradoxes) in the contemporary scholarly communication system that relate to the linguistic representation. We do not approach the matter as a question of equal representation; this is not feasible, reasonable and in practice sustainable. However, it is a question of equal chances in having the vital space to communicate and to contribute both to science and society, without adherence to standards that have a few side effects, including the cultural coherence of a country.
Framework
Why is multilingualism so important? In a world where internationalization has pivotal importance, we question the very essence of this process; is speaking one single language the way to internationalize? Elea Giménez Toledo [2024] highlights that “…research has an economic, social and cultural impact on its surrounding environment, as when scientific knowledge is disseminated it filters through to non-academic professionals, thus creating a broader culture of knowledge sharing.” Therefore, multilingualism in science, as a practice that promotes equity, is a mechanism for economic, social and cultural cohesion and advancement.
It is most obvious that the internationalization of scholarly communication is a considerable factor to advance scientific collaboration and to expedite the solution of global research problems. We can acknowledge that English is an effective conductor of communication and their use is widespread in many fields, not just in science. This is reflected in the statistics between the number of native English speakers and the actual speakers in the world (from 380 million of native speakers to approximately 1.5 billion speakers, see Ethnologue, 2024). But, in the world of science this wide-spread use of English has distorting side-effects.
Sociopolitical advancements are critical for the so-called hegemony of the English language and the consolidation of the scholarly communication is not attributed just to language, but also to the platforms that shape decision making on certain levels. We should remember that the prevailing tools for decision making in science and academia are platforms and portfolios of services that are owned by a few commercial colossuses.
Apparently, there are historic roots for these practices, including the origins of these databases and the rationale behind their development. The production of these tools in the past had several limitations, which lead to this high selectiveness, with indexing of cited references being one of them. While the technology has rapidly evolved and these colossuses claim that have invested on their systems, this has not happened to the direction of representing anything else than it was represented forty or seventy years ago, which, as a fact alone, leads to the conclusion that this is a deliberate political decision.
Problems are evident to the authors, the journal editors, the specific research communities, the policy makers, the libraries, etc. They start as early as the authoring of a research publication when the only way to be visible is to write in English. Ramírez-Castañeda [2024] provides a summary of the challenges that the researchers in biological sciences (which cannot be considered as a neglected field) face, including the extra labor, the costs, the probable biases, etc. Other socioeconomic barriers can be listed, such as the cost to access services, either for publication (APCs), or for language editing services, which is a prohibiting factor for many researchers in the world. Researchers quite often face policies that do not promote linguistic plurality, but rather limit what participates in the scientific dialogue, see for instance the cited literature. This might be less frequent in Open Access journals, but Arenas-Castro and his colleagues [2024] mention that “Although most editors-in-chief indicated that they allow or encourage citing non-English-language literature, only 10% of the journals explicitly mentioned this in the author guidelines…”
The main problem for journal editors and the communities that these journals represent is the high threshold for entrance in indexing services. Low journal visibility risks the time and effort that the researchers invest. Consequently, this jeopardizes the viability of research communities that, for their own reasons, prefer to publish in specific journals. Publishing in those venues has its own merit in studying specific fields. Kraus [2022] mentions about the role of monoligualism in studying political sciences that “There is a significant overlap between linguistic cultures and political cultures. By extension, this overlap has consequences for how we do political science, as particular political concepts are linked to particular linguistic cultures”. Furthermore, this underestimates their contribution and undermines their position in the formal assessment processes. Gradually, this has other, less formal, effects, such as the low social recognition and esteem. This is strongly associated with specific fields, such as Social Sciences and Humanities. Yet, these effects are critical, because these fields define to a large extent the understanding of unique issues, problems and concerns in a specific country. Furthermore, areas such as history, archaeology or philosophy, have a long-standing tradition in the cultural identity of a country.
Therefore, researchers invest more resources in preparing their works, in preparing alternate/translated versions of these works (see OPERAS White Paper on Multilingualism on the perceptions regarding translations), in securing funding -more than the purchasing power of their country allows- and at the end they risk of not being attributed to of their effort (most often in their own country).
Method
In this study, we use data from the Directory of Open Access Journals, which are often compared to data from Scopus. The data have been mined in April 2024, they have been processed in Google LookerStudio, after cleaning and homogenization, and have been visualized in Flourish. The DOAJ data were preferred for a number of reasons. While fewer in numbers than the data in other indexes, we have evidence that they hold information for more languages than the rest (see below). The associated metadata are equally, if not more, rich and well-structured to those in other databases. The study takes as a stable parameter the concept of small languages, meaning that we are concentrating on national languages. Other problems are present when one studies minority or endangered languages. For instance, Grange [2024] mentions certain main problems for journals in minority languages, namely additional work and scarce means to publish and neglect from the design and access from scholarly communication infrastructures or from funding the same operational costs. These are out of the scope of this study.
Findings
Lingua… Greca
We often call English as the Lingua Franca of scholarly communication. It did not happen overnight. So, the start was to look at extent of the Greek journals inside the country’s context and how they relate to other languages and especially English. To do so, we checked two sources: first, the holdings of the Greek libraries as recorded in the Greek Union Catalogue of Periodicals, which is maintained by the National Documentation Center. There are ~4.000 journal titles in Greek, which correspond to 4,65% of the overall holdings. Obviously, there is a lot of historic information; it is difficult to determine which of these journals are still active. Historic information can be also found in the two ISSN registry services. The database of the National Library of Greece provides the information that 10.500 ISSNs have been provided to periodical publications. The portal of the ISSN International Center provides an approximate number, ~10.200 journals, out of which ~1.000 are electronic and only 110 are labeled as Open Access. Again, these data should be treated carefully; not only they contain journals that are not active anymore, but also periodical publications for any kind of audience and in any format, such as newspapers, ephemera, public journals, society journals, professional publications, etc.
Journals and Speaking languages
Articles
The next step was to estimate the number of the articles that are published in English. We found that they represent 76,6% of the articles in DOAJ journals that have English are their first language. The second language is Spanish, which corresponds to ten times less (7,9%). In order to reach to a more accurate estimation, we saw that for the journals that publish in only one language and use English, then this percentage climbs to 85,7%.
Origins
Topics
The next check was about the topics per language. Most of the medical journals are being published in English, followed by general Science, Technology and Social Sciences journals. There is also substantial representation of the topics of Philosophy and Agriculture in English, while the journals in Spanish prefer the topics of Social Sciences, Education and Language & Literature (a similar pattern also found in Portuguese). Journals in Indonesian, a market that has been boosted the last few years, mostly focus on Education and Social Sciences.
Indexing
We also checked how many of these journals are visible in Scopus. Except for journals with Catalan as first language (representing though a small fraction), which are more than 50% present both in DOAJ and in Scopus, all the other journals are listed less than 50% in the latter database. Journals with English as first language are 48,9% present, followed by Chinese with 42,6%. Journals with primary languages, such as Arabic, Persian and Indonesian, are present in both services to 7,52, 8,55 and 4,9% respectively. However, when limiting the sample to the journals that are published in UK and USA, two natively English-speaking countries, this number reaches 70% of presence in Scopus.
Paradoxes
The first paradox is that despite the advancement of strong statements and policies that promote multilingualism and Open Science, such as the Helsinki Initiative for Multilingualism in Scholarly Communication in 2019, the growth of journals is governed by titles in English. We can see an increasing trend; from 2016 to 2022, the journals that publish in one language are publishing mostly in English (from 70,1% of the new titles to 89,8%).
The second paradox is that of academic freedom. How can one be free to conduct research on their preferred topics when the passport of visibility requires authoring in English, as well as staying relevant to an international audience? It appears that the use of research intelligence tools suggests areas that satisfy the international readers. We are using an example from Philosophy, a research field that is linked with the Greek context.
While preparing a report for one of our Departments, we processed data from SciVal, the research intelligence tool of Elsevier, that showed for the last five years the production of the University on the topic of Philosophy was on (ranked by size): (1) Aristotle; Plato; Socrates, (2) Kant; Theory; Epistemic, (3) Science; Risks; Nanotechnology, (4) Martin Heidegger; Nietzsche; Phenomenology and (5) Triangle; Proof; Mathematics. At the same time, the topics according to the Topic Prominence metric, “an indicator of the momentum/movement or visibility of a particular Topic” [SciVal, 2024], were (1) Science; Risks; Nanotechnology, (2) Kant; Theory; Epistemic, (3) Justice; Theory; Human Rights, (4) Rhetoric; Feminist; Feminism, and (5) Martin Heidegger; Nietzsche; Phenomenology. If one wants to take a decision, that in the Humanities might be an existential one, then writing in English and on topics that satisfy the international aspect is the only option that can be considered.
Flower on the Sandy Beach
This study was based on the evidence that the English language governs the scholarly communication. Using the information of a community owned database, we reached to some insights that lead to the observation of a few distortions and paradoxes. It is not within the scope of this study to “condemn” English and commercial tools, another critical factor, for this consolidation. What was intended was to provide the evidence of the magnitude of the phenomenon and to ask to consider for solutions that provide equal chances to researchers. Those who are being asked are both international and national stakeholders, from academics to policy makers and from publishers to researchers.
Policy makers, in academia and research, should reconsider the weight of the national languages in the research assessment. These not only sustain the inequalities of the international system, but further do not address the local social, economic and environmental challenges. The narrative of internationalization should be revised for countries that have a long-standing history and their contributions over the centuries should not be oppressed by contemporary practices and tools. Decentralizing the commercial tools from the decision-making processes and considering community owned services can be a valid option. Community owned infrastructures tend to respect national characteristics; the example of other research assets, such as the data resources in CLARIN’s Virtual Language Observatory, is indicative.
Secondly, the Open Science field must highlight these inequalities and advocate for the existing best practices that work for the visibility chances of all researchers, regardless of origin, field and language. It must help to build the determination to take the bold, but reasonable decision that alternate systems must exist and that there is no room for one-size-fits-all approaches. As a folklore Greek song says, “A fish does not live on land, nor a flower on a sandy beach” (“Στη στεριά δε ζει το ψάρι ούτε ανθός στην αμμουδιά”). In the same sense, some fields cannot be evaluated in the environment and conditions of another paradigm.
To return to the Greek context, the exposure of Greek journals (published in Greece), even in DOAJ, is minimal. There are 44 journals in DOAJ with just two of them accepting papers only in Greek, one in Medicine and one in Social Sciences. In Scopus, the journals that accept papers in Greek are 23 (note: there are also listed 12 book series), mainly in Social Sciences and Medicine. There is a glimmer of hope, because they are quite systematic in exposing their content, as 49% (n=19) are listed both in DOAJ and Scopus. However, only two of them have the DOAJ Seal, which indicates their commitment to high quality services.
Combined with observations from the national environment, this leads to the natural conclusion that the whole sector is dependent on the format and the majority of journals are either print ones, or not properly exposed. Therefore, it seems to be an awkwardness of what the transition to the digital environment entails. The digital environment is often associated with quick results, but this can be true up to the point of registering a journal. However, its potential impact can be revealed after many years and often the editorial staff and administrators do not provide this timeframe. It is common place that authors in Social Sciences and Humanities are in favour of the printed medium, which is also something that the Greek publishing sector exercises more. There are no business models, not just in the commercial, but in the academic sector as well. Certain logistic processes, specifically for those journals that are published by academic units and libraries, do not encourage the flow of income for journals to sustain their operation. Therefore, this marks a responsibility for publishing and editorial entities. The transition should be executed in a more professional way, taking into advantage the tools that the community of Open Access has developed over the years.
While there is a set of services available to some publishing entities, these are not coherent and embedded in a national policy. This means that these journals are not visible even in their own country. At the same time, the national policies regarding academic assessment and promotion adhere to certain STEM paradigms. The Hellenic Authority for Higher Education is only accepting contributions that are indexed in Scopus, meaning that other journals, books, reports, etc. are ignored.
In conclusion, the Open Science practices and principles can be used to create the conditions that provide equal chances for scholars around the world to be properly appreciated and to contribute to their own countries’ cohesion and growth.