The conference focusses on three distinct but related topics relevant in the assessment of SSH research: the importance of books; the importance of plurilingualism/multingualism; the importance of integration/impact.

LERU has been asked to specifically comment on issues in the evaluation of multilingual publications. What follows is not a structured position paper but rather a deurmekaar ¹ sequence of comments variously relevant to the topic.

We understand “multilingual” in a very broad sense, designating separate publications by the same scholar(s) in different languages, publications with different languages in one text (e.g. English-language discussions of French philosophers containing quotations in French and German, possibly accompanied by translations), edited volumes or journals containing separate texts in different languages. . . In addition, multilingualism is also an important feature of the wider research ecosystem (so not just publications), involving procedures in grant applications and assessments, doctoral training and undergraduate research-based teaching etc...

Let’s call a cat a cat (et Rolet un fripon. . . ): The English elephant in the conference room

The conference programme interestingly avoids mentioning that what is arguably most visibly at stake in this context is the impact of anglobalisation. While a lingua franca, like English is today in the world of scholarly communication, obviously has many advantages, it can also have unhelpfully homogenizing effects, stifling not only linguistic but also cultural and indeed scholarly diversity—this latter both in terms of alternative (complementary or competing) approaches to specific issues and in terms of academic policy and protocols, including publishing traditions and assessment practices.

It is perhaps worth recalling that the very term lingua franca (as Wikipedia, the prime vehicle of intellectual-anglobalisation-in-the-nicest-possible-way, tells us) “is taken from the medieval Mediterranean Lingua Franca, an Italian-based pidgin language used especially by traders in the Mediterranean Basin from the 11th to the 19th centuries”, suggesting a medium facilitating the exchange of quantifiable commodities rather than contributions to communal reflection.² That academic communication today indeed has a dominant lingua franca in the first place is therefore arguably closely connected to the dominance of quantification-based disciplines in academia. It could have been French, German or (perish the thought) Dutch—the point is it’s upgraded pidgin.

[An aside: “Pidgin” itself, of course, is a loaded term, being “Pidgin” Chinese for “business”³, which opens up a Pandora’s box of legitimate post- and de-colonial qualms and concerns which we cannot adequately address here. Just one example of the complications this can bring to academic practice, specifically in terms of academic teaching: Students now have access to unimaginable amounts of publications through the research infrastructures of our university libraries. Doing research for their papers on, say, a novel by Thomas Hardy or Aldous Huxley, undergraduate students key in the title of the novel and instantly receive a substantial set of results, many of which are published in predatory or semi-predatory journals publishing material by authors who are often not native speakers and who have had to pay extortionate article processing charges. The quality of thought in these articles varies, as it does in all articles, but the English is often idiosyncratic. For students whose native language is not English, it is not always clear that this may not be the English they should seek to master. For their teachers, it is hard to just dismiss these sources on grounds of linguistic infelicity because they often do contain valuable thinking—and indeed thinking that may be refreshingly critical of the dominant anglobal discursive regime. To be continued.]

In what follows we focus on the impact of the dominance of the English language on academic work in Social Sciences, Humanities and the Arts, but it is important to bear in mind that the natural and biomedical sciences are also adversely affected by this imbalance.⁴

Impoverishment of local republics of letters

The dominance of a lingua franca can have an impoverishing effect on the discursive ecosystem of countries or regions with local languages different from the lingua franca, especially when performance-based research funding systems are in operation. A Spanish social scientist investigating rural precarity, for instance, may prefer to publish in leading journals in their discipline, which will tend to be anglophone, rather than share their findings with the wider audience in Spanish publications that may be valued less or even ignored altogether by funding bodies and performance assessment committees. ⁵

An important further observation in this regard is made in a recent study co-authored by (among others) four of our colleagues present here (Emanuel Kulczycki, Janne Pölönen, Tim Engels and Gunnar Sivertsen). More than 66% of SSH researchers at Flemish universities, they find, only publish in English, not in a local or indeed in any other language ⁶. It would appear then that two thirds of Flemish SSH academics barely engage with the republic of letters in the community that hosts them and in most cases also pays their salaries. Put differently: the tendency to prioritise publications in English in the assessment of researchers may have the unintended side-effect that researchers are discouraged from engaging with the wider local public. Causal connections are of course hard to establish here, and there may be bias in the databases analysed in the study by Kulczycki et al. (as the authors indeed acknowledge), but the data are arresting enough to warrant caution.

The relatively recent trend in researchers’ assessment to explicitly value various forms of public engagement and outreach constitutes a salutary corrective here, provided it is more than mere lip service. As Gunnar Sivertsen has argued, what is required is a sustained commitment to “balanced multilingualism as a basis for governing the tensions between strategies for internationalization and excellence in research on the one hand and strategies for societal relevance and participation on the other.”⁷ A relatively recently developed research assessment tool in the Netherlands sets an inspiring example in this regard. QRiH (Quality and Relevance in the Humanities) explicitly recognises so-called “hybrid publications” aimed at peers but also at the wider public, and therefore often published in Dutch.⁸ The recognition of such hybrid publications in the local language as valid research output can only be applauded.

Evidently, for a balanced multilingualism model to work, the existence of a sustainable infrastructure for publishing in local languages is of the essence.

Peer review problems

The dominance of English as the lingua franca for scholarly communication is not only a matter of language but also involves publication practices, including a widespread commitment to (typically external double-blind) peer review (PR). This is problematic not only because PR in itself has multiple possible drawbacks, but because setting it as the standard requirement for scholarly communication discriminates against alternative editorial traditions in other-than-English contexts that may be just as valid. Editorial decisions in many German and French journals, book series and publishing houses, for instance, are made at the discretion of an often fairly select board of editors who do not or not always subject submitted manuscripts to external PR but primarily rely on their own judgement, typically with reference to a track record of decades in the business of publishing quality work.

As one particular instance of the imbalances this may generate, consider the case of the Flemish Academic Bibliographic Database for the Social Sciences and Humanities (VABB-SHW), an annually updated list of publications by SSH scholars employed at Flemish universities used as part of the government’s performance-based research funding system.⁹ The principal rationale for VABB-SHW is that it allows the government’s performance radar to pick up academic publication output that does not appear in Web of Science, as is the case for much if not most of the work in SSH. Prior to the implementation of VABB-SHW, the vast majority of scholarly output in SSH just did not make any difference for the universities’ research budgets; now SSH-output not in Web of Science does make a small but not negligible and in fact increasing difference.

The operational principles of VABB-SHW are fairly straightforward: Every year, each of the Flemish universities pumps the publications of their staff into a system which then separates the SSH sheep from the Web of Science wheat, and herds the SSH sheep to the grandly named “Authoritative Panel”. The panel then runs a check on the input based on a set of minimal criteria, the most important of which is external peer review. Anything published in journals or by publishers that do not openly advertise an external peer review policy is kicked out; everything else counts towards research budget for the universities.

An important drawback of VABB-SHW is that it is structurally detrimental to balanced multilingualism. For while the system does at least acknowledge that output not picked up by run-for-profit providers of hot-air analytics like Thomson Reuters and now Clarivate may well be worthy of respect, it fails to pick up many high-quality publications that do not meet the requirements of external PR strictly understood, and that includes many publications in languages other than English. As a recent analysis shows, of the total of publications registered in VABB-SHW about 80% are in English, and only some 16% are in Dutch, the main language in Flanders. More surprising perhaps than the massive predominance of English is that publications in French, the second national language in Belgium, only add up to less than 3% of the total. ¹⁰ Obviously, many more excellent scholarly publications in French are authored by academics in Flanders, especially by scholars working on French history, literature and culture, but since they are published (mainly in France but also in Belgium) by journals or presses that do not disclose editorial protocols compliant with external PR requirements, this scholarly output is ignored. And even though this should in principle not matter much to individual researchers if VABB is just used as the mindless aggregate budget allocation instrument it is designed to be, it still causes understandable displeasure in the scholarly community.

The bottom line is that criteria like external peer review for the quality assessment of publications are not language-neutral: it is never just a matter of publishing in a different language, there is always a language-related publication culture that has every right to be respected. And this quite apart, again, from any misgivings one might rightly have about peer review as such.

SSH disciplines with forum languages other than English

SSH contains many disciplines engaging specifically with material closely connected to local, regional or national contexts and cultures, which inevitably entails the necessity to engage with and indeed to use the language(s) of those contexts and cultures. Such a language particularly pertinent to a discipline is sometimes called the “forum language” of that discipline. Examples of disciplines whose primary forum language may not be English include, obviously, the study of separate languages and literatures, but also area studies, subdisciplines of history, law, politics, or public governance devoted to specific regions etc.

In some disciplines, publishing in languages other than English continues not so much because of the specific nature of the material being studied but as a result of the history of the discipline itself. Papyrologists, for instance, continue to publish in German or French or Italian because those were the leading languages when the discipline was historically designed.

Other disciplines are also expected to publish their findings for a more professional audience who may not have the English language skills required. Archeologists in Barcelona, for instance, routinely publish in Catalan for the benefit of readers working in the local archeological community.

When judged by peers in the same disciplines, scholarly output in these disciplines in their forum language is likely to be assessed fairly. However, when scholars with a strong track record in these disciplines are assessed by general review bodies—for instance in interdisciplinary panels, or when they submit more general theoretical proposals in English—they risk being ranked less favourably, which effectively sends them packing, back to their own choir.

The cold comfort of translation

In branches of a number of SSH disciplines such as philosophy, theoretical sociology, literary theory and theology, the primary “data” are often extremely complex texts demanding almost pathologically close reading. These texts comprise not only the corpus of the canon but also critical commentaries or more general academic reflections, and they can be written in different languages. Adequate research on these data requires enhanced multilingual reading skills, yet such skills may be less common than they once were.

Some three decades ago, students of these texts in many European countries would therefore typically be academically multilingual. A Belgian PhD student with Dutch as their native language interested in post-structuralist approaches to English literature, for instance, would be expected not only to have near-native English but also to sufficiently master French and German to read, say, Lacan, Blanchot, Derrida, Deleuze, Nietzsche, Heidegger, Adorno and Benjamin in the original, and they would typically quote these authors in the original in their English-language publications. English translations of these authors were available, of course, many of them excellent, but close reading will always encounter passages where translational decisions add or (more often) subtract twists in thought, making a return to the original rewarding if not therefore necessarily conducive to conclusive understanding. To native speakers of English, such display of familiarity with the original in order to defamiliarise the translation sometimes came across as pretentious and pedantic—which it probably often was. However this may be, by and large those days are gone. Major trends in cultural humanities are now overwhelmingly English-only. Work in other languages is picked up almost exclusively in translation. Yet the comfort zone of translation can also be a place of intellectual deprivation and problem loss.

APC & BPC issues

Article processing charges (APC) and book processing changes (BPC) also lead to imbalances in the scholarly communication ecosystem. Publishing an article in a local publication in a local language is typically free of charge for the author, but even though the journal may well observe a strict PR protocol the publication risks being valued less than a publication in an anglophone international journal charging often substantial APCs. This especially disadvantages junior researchers who do not have access to a research budget to cover these costs.

On the other hand, a scholar seeking to publish a book in German with a German publisher is often expected to pay thousands of euros for the privilege. As an especially disturbing instance, a German and a Belgian scholar who recently completed a commissioned 5-volume scholarly edition of Gesammelte Schriften by an important German-Jewish female intellectual were informed that the publisher, who had initially confirmed that funding was secured, now expects them to fork out 50.000 euros themselves.