Developing research systems that promote diverse, multilingual and relevant research for different audiences is a key and often overlooked element in making research accessible. However, biases in traditional research assessment often place researchers looking to produce multilingual research outputs at a disadvantage. Reflecting on the European Commission’s recently published aims for the reform of research assessment Janne Pölönen, Emanuel Kulczycki, Henriikka Mustajoki and Vidar Røeggen, suggest the omission of multilingualism from this agenda, risks undermining the project’s aims of supporting high-quality and accessible research.
It is important to communicate research results to international audiences. However, when research is communicated exclusively in English, it is inherently limited. This was apparent during the COVID-19 pandemic, in Europe and Globally, which saw a widespread need for multilingual scholarly communication, not only between researchers, but to enable research to reach decision-makers, professionals and citizens. In recent years, policies for Responsible Research and Innovation (RRI) and open science have called for increasing access to research and greater interaction between science and society. This can only be possible if research is communicated in multiple languages, including those actually written and spoken locally.
A new report published this week, resulting from the European Commission consultation with international and national stakeholders, paves the way to a European agreement towards a reform of the research assessment system to be concluded in Paris in February 2022. Instead of narrow focus on research, publications and metrics, the report outlines assessments rewarding a broad range of open science practices, recognising diverse outputs, activities and impacts of academic work, as well as respecting differences between fields. Overall, the report is very welcome, addressing as it does key areas of reform that have emerged over the past decade and have been raised in a range of responsible assessment declarations, policies and agendas. Multilingualism, however, is the most notable omission.
Multilingualism, the neglected dimension of research assessment
It is increasingly recognised that “multilingualism is an important but often neglected dimension of diversity in research, helping to ensure that research remains locally relevant and accessible”. Starting with the DORA declaration in 2012, responsible assessment entails valuing diversity of outputs, activities and impacts. The Leiden Manifesto on research metrics specifically recommends protecting excellence in locally relevant research and warns against equating excellence with English language publications in high impact factor journals indexed in the Web of Science.
Making this link between multilingualism and assessment clear, was also one of the main goals of the Helsinki Initiative on Multilingualism in Scholarly Communication, which aims to promote language diversity in research assessment, evaluation, and funding systems. The Helsinki Initiative recognises that language biases are produced in both evaluations based on research metrics, as well as evaluation based on expert assessment, as stated in two of the initiative’s goals:
- Making sure that in the process of expert-based evaluation, high quality research is valued regardless of the publishing language or publication channel.
- Making sure that when metrics-based systems are utilised, journal and book publications in all languages are adequately taken into account.
Language biases in research assessment
Perhaps the clearest example of everyday language bias comes from the manuscript peer review, where it is a frequently reported experience of non-native English speakers that reviewers judge their research based on the quality of their writing instead of the content. But, in all kinds of expert evaluation procedures, the selection of assessment language can discriminate against researchers who are not native speakers or fully fluent in the given language(s). Consequently, the selection of evaluators, and their language skill, should be a relevant concern in assessments.
One challenge in assessment is that excellence is too often equated with English language publications, especially those published in journals having an Impact Factor or those indexed in Web of Science and Scopus databases. While Journal Impact Factor is a major concern for responsible assessment, a European Commission report published already in 2010 identified the limited coverage of research output in Web of Science and Scopus databases as the key limitation in assessment of European university-based research in major academic fields, such as the Humanities, Social Sciences and Engineering.
in all kinds of expert evaluation procedures, the selection of assessment language can discriminate against researchers who are not native speakers or fully fluent in the given language(s)
Comprehensive publication data from Finland shows that more than 70% of non-peer-reviewed outputs – including publications targeted at professional and general audiences – are in Finland’s national languages (Finnish and Swedish) in both STEM and SSH fields. In addition, SSH researchers publish a large share of their original peer-reviewed research in national and regional fora, and in languages other than English. Evidence from seven European countries shows that Web of Science and Scopus databases cover only 25–31% of the 164,218 peer-reviewed journal articles published by SSH researchers in 2013–2015, and only 3–8% of the articles published in the local languages of the seven countries. Thus, intended or unintended language priorities in assessment may lead, for example, to systemic undervaluation of SSH research and researchers.
Multilingualism as a European challenge
According to the joint statement of the European University Association (EUA), Universities Finland (UNIFI) and Universities Norway (UHR), “Multilingualism is particularly relevant for Europe, as its research is characterized by geographic, cultural and linguistic diversity and the common principle of excellence”. Indeed, the EU has 24 “official and working” languages. According to the Europeans and their languages report published in 2012, almost half of the Europeans (46%) were not able to speak any foreign language well enough to hold a conversation, and only 38% were able to do so in English (the most widely spoken foreign language).
Ideally, language is a non-issue in assessment. Researchers should be recognised and rewarded according to the results and the impact of their research
Multilingualism and language diversity are also a matter of rights. The United Nations’ Universal Declaration of Human Rights article 27 states that “everyone has the right freely to participate in the cultural life of the community, to enjoy the arts and to share in scientific advancement and its benefits”. The Charter of Fundamental Rights of the EU also places an obligation on the Union to respect linguistic diversity (Article 22) and prohibits discrimination on grounds of language (Article 21). The European Commission has a strong multilingualism policy for promoting language learning and linguistic diversity across Europe. However this policy has not addressed the value of multilingualism in science, research and innovation.
According to the recently approved Unesco Recommendation, one of the main goals of Open Science is to make “multilingual scientific knowledge openly available, accessible and reusable for everyone”. It recommends that the member states consider “encouraging multilingualism in the practice of science, in scientific publications and in academic communications”. Ideally, language is a non-issue in assessment. Researchers should be recognised and rewarded according to the results and the impact of their research, despite the language of application or publication. In practice, assessment criteria and methods are often far from language-neutral.
If any of these initiatives and obligations to promote an equitable, diverse and inclusive academia, and make available multilingual knowledge for all sectors of society are to be achieved, European research assessment reform needs to address language biases in recognition and rewards.
Note: This article gives the views of the author, and not the position of the LSE Impact Blog, nor of the London School of Economics. Please review our comments policy if you have any concerns on posting a comment below.
Image Credit: Elf-Moondance via Pixabay.
A very interesting article and an important topic. I realise that as a native English speaker with only mediocre Norwegian, my opinion may come across as hopelessly biased and self-serving, but here goes: I recognise the importance of communicating science in a variety of languages to a public that may only speak their local language. I also recognise the importance of being able to publish in a local language research that has only local relevance (although one has to wonder how that would be established). However, I worry that reforming the incentive system in this way would be better for many individual researchers but harmful for the progress of science as a whole. As described by Michael Gordin in his excellent book Scientific Babel, a scientist looking to keep up-to-date in their field 100 years ago would have needed to speak a mix of German, English, and French, while a chemist in the 1970’s would have really benefited from an ability to read Russian. In other fields, Japanese. For historical reasons, English has become nearly universal – far more ubiquitous than Latin ever was – as the language of the peer-reviewed literature. Yes, that is a result of an Anglo-American imperialist hegemony of which I am only too aware (and of which I often despair) but it does make it far easier for the billion people that have English as a first or second language to be involved in the scientific endeavour. It certainly makes it easier to find peer-reviewers. The current ubiquity of English is unprecedented, and arguably it is a good thing. Does the sort of reform that you are advocating not risk a return to linguistic – and potentially nationalistic – silos?