We can talk in English, but can we talk about English?
By Basile Boulay (part 2 of 3)
In this second of three blogposts on the linguistic predominance of English in research (first one here, third here), I explore the variety of ‘everyday’ barriers faced by researchers whose first language is not English when producing, presenting and/or publishing their findings in this language. The focus here is on the immediate issues they face in the current context, and potential solutions. The deeper considerations of political economy and power will be the subject of the third, final blogpost.
Literature on this topic remains fairly scarce. References therefore include fields other than Development Studies, as there’s no reason to believe that challenges faced by development researchers are fundamentally different. Similarly, our reflection about Development Studies also applies to other fields beyond disciplinary boundaries. Without negating the specificities of Development Studies, research under internationalised and marketized academia imposes its trends on all disciplines.
More time, more stress, more money
A recent and useful recap of the barriers faced by non-native English researchers is provided in a study aptly titled: The manifold costs of being a non-native English speaker in science. Focusing on the field of environmental science, it presents results from a survey with 908 researchers to establish the extent of linguistic discrimination. It constitutes a useful starting point to understand the current situation, the usual quantitative limitations of such exercises apart.
Five key categories of scientific activities are considered: paper reading, writing, publication, dissemination, and presentation in conferences. The results are unambiguous: non-native English speakers perform ‘worse’ than their native peers in all dimensions, and the effect is particularly strong for early-career researchers. Non-native English speakers need more time to read and write in English and are more often required to use language proofreading services. Once articles are submitted, the likelihood to be rejected by journal editors is also significantly higher: ‘the frequency of language-related paper rejection is 2.5 to 2.6 times higher for non-native speakers’ (p. 5). Barriers are also important when it comes to disseminating the work produced, particularly at international conferences: ‘about half of the early-career non-native English speakers of high-income nationalities (Japanese and Spanish combined) often or always avoid oral presentations due to language barriers’ (p.6). In addition, non-native speakers need an estimated 94% extra time compared to their native peers to prepare/practice for conferences.
Although the authors of the study stress that the exact quantitative estimates may not apply to other fields, anecdotal evidence suggests that these biases also exist in social sciences and Development Studies. Faraldo-Cabana reports similar patterns are reported by for the field of social sciences and criminology: ‘Nonnative researchers are at an obvious disadvantage to native speakers regarding the lexico-grammatical, rhetorical and stylistic use of the language. This means difficulty with producing written English at an acceptable level and can lead to editorial rejection’ (p.169).
When it comes to conferences, the EADI secretariat often receives requests from potential participants to present their work in other languages, due to the inability or discomfort of presenting papers in English. This was obvious at the 2023 general conference in Lisbon, and one of the reasons why some sessions were organised in Portuguese. It is important to appreciate these concerns since they are particularly hard to quantify or estimate. Added stress (and in the worst cases, anxiety) when presenting at conferences is an additional ‘cost’ borne by non-native speakers.
Another problem worth pointing out has to do with financial resources. Publishing in English does not only require a lot more time, it also often comes with a need for full text revision using professional services, which charge significant fees. This further deepens the gap between native and non-native researchers, particularly when the latter are in institutions in the Global South with less financial capacity than their Northern counterparts.
What can be done?
The problems mentioned above suggest ‘unfair access’ to research outlets based on linguistic discrimination and the gatekeeping mechanism effectively enforced by the linguistic predominance of English. For this reason, I name the potential solutions to these issues fist-level, or instrumental responses, as opposed to second-level responses which are concerned with deeper questions of power, as I will explain in the following blogpost. These first-level solutions focus on facilitating publication in English for non-native researchers. Broadly, they can be boxed into three categories: time, financial resources, and technology.
When it comes to saving time, specific initiatives can include, for instance, specialised training on producing a crisp abstract in English, or improving the quality of introductions before the first submission to a journal. This usually entails some kind of mentor/mentee relationship, which requires in turn a commitment from mentors to dedicate some of their time to mentees for free, an aspect also emphasized by Amano et al.
Some of the existing initiatives can also be sources of inspiration. For example, the European Journal of Development Research has a special editor focusing on submissions from early-career authors to provide specific advice. A similar role dedicated to language issues could be envisaged. Publishers themselves could also provide such help, offering advice on maximising chances for papers to make it to the review stage and ultimately to publication. Together with a clear institutional commitment by the party offering help to non-native speakers, such direct help is a way to rebalance the unfair situation from which Global North English-speaking institutions benefit at the expense of non-native speakers. Therefore, an argument in favour of fairness and justice is to be made. Time is valuable, and its transfer in the form of help with language is an important lever. A similar time commitment could be expected at international conferences, where chairs could be asked to spend more time preparing feedback to non-native speakers.
The second lever is financial, for example by contributing to the proof-reading services required by journal editors/referees for a paper to be considered for publication. The question of course is: who should finance that? While there is no easy answer, the question of fairness is again important. Opening a collective reflection on this issue could yield interesting perspectives, and potentially unexpected ones. For instance, could the publishers themselves be asked to financially contribute to the fee for proof-reading? This may at first sound like heresy to some readers: why should publishers finance proof-reading? It should be noted, however, that academic publishers have been making enormous profits over the past years, with the profit margin of Elsevier and Springer Nature at around 40%, as explained in a recent article from Alternatives Economiques. These profits are based on largely publicly funded research, as well as peer-reviews done free of charge. So, while universities use (mostly) public money to fund research, they must also subscribe to publishers’ services to access review and journals or finance open access articles. An example cited in the above-mentioned article: French universities spent a total of 87.5 million euro in 2020 to subscribe to journal access. In this context: is it really too much to ask publishers to routinely offer or finance proof-reading services, at least partly? Again, our collective reflection on this must not omit a central aspect of neoliberal academia: the public sector is the main funder of it, the private sector the main financial beneficiary.
The third lever is technology, with the growing possibilities offered by artificial intelligence tools. It is still unclear to which extent linguistic assistance tools will develop in the coming years, but it is already the case that some tools can improve writing and reorganise texts to enhance readability. ChatGPT, for example, is already able to do that. However, this may well correlate with financial capacity, as free versions of AI tools tend to be limited and may not offer as good a help as premium ones. Here too, then, the issue of financial limitation comes into play, and some financial transfers could be envisaged. For example, specific institutions in the Global North (universities, but also publishers as mentioned above), could finance subscriptions to AI tools for institutions in the Global South, particularly in non-English speaking countries.
Instrumental … but also political
While the above deliberations are largely instrumental, i.e. they only address the immediate situation of unfairness, we nonetheless quickly fall back on political considerations when it comes to debating who should be funding these responses, be it in terms of time or money, which is perhaps why change is so hard to implement. At the end of the day, it is a matter of justice, as summarised here (p.10): ‘To date, the task of overcoming language barriers has largely been left to non-native English speakers’ efforts and their investment in ways of improving their English skills’. And if you wonder what is the proportion of the world’s population which is non-English native speaker, the answer is: around 95%.
Basile Boulay is Senior Executive at the EADI secretariat and holds a PhD in Development Economics
Image: Mathieu Stern on Unsplash