English and Academia: Access, Publication and Unequal Costs

English opens doors in research. It also puts a price on entry.

A researcher can have a sound argument and still struggle to make it heard in English. The practical response is to help that researcher communicate. The institutional response should include asking why the extra work falls so unevenly. Both responses are necessary.

For a researcher preparing an English-language paper or conference presentation, the immediate task is concrete: make the question, evidence and limits understandable to another specialist. Fluency helps, but it cannot repair a weak research design. Equally, an awkward sentence does not invalidate a strong one. Academic language support should keep those distinctions visible.

My argument here is that English teaching for researchers works best when it follows the work they need to do. “Improve your English” is an intimidatingly large assignment. “Explain why this comparison supports your claim” gives us somewhere to begin.

Reading: decide what you need to find

Begin a literature search with a question and a record of your search terms. Use titles and abstracts to identify likely relevance, then inspect the methods, results and limitations that matter to the question. Reading every paper from the first word to the last is a method of being busy; it is not automatically a method of being well informed.

A useful reading note separates three things: what the authors did, what they found, and what they infer. Add a fourth line for what you still need to check. Translation tools can help you work through a difficult passage, but compare crucial claims with the original, especially when a negative, a qualification or a technical term changes the meaning.

English searches also need an explicit boundary. If the review concerns a place or practice with relevant literature in other languages, decide how that literature will be sought and assessed. Otherwise, the search language can quietly become an exclusion rule.

Publication: make the claim inspectable

Before polishing the prose, ask whether a reader can identify the research question, the comparison, the evidence and the conclusion. A paragraph can be grammatically elegant while concealing a missing step in the argument.

Consider this invented sentence: “Our programme improved students’ confidence.” If the study collected only a survey after the programme, that wording asks the evidence to do too much. A narrower statement might be: “Participants reported confidence in the survey administered after the programme.” It still needs details about who responded and what was asked. It no longer claims to have measured improvement or established its cause.

Language editing is valuable partly because it exposes these decisions. Co-authors should agree on the intended claim before an editor or a tool makes it sound more convincing.

Conference preparation needs an unscripted part

A prepared talk lets a researcher control the order of the explanation. Questions remove that comfort. Rehearse with someone who asks for a definition, challenges a comparison, or proposes a different interpretation.

I would prepare thirty-second, two-minute and five-minute explanations of the same study. Each version should preserve the question, principal finding and limitation. Extra time adds context. It should not be necessary to wait five minutes to discover what the study found.

Then ask the listener to restate the claim. If their version is stronger than the evidence permits, revise the explanation. Useful responses include “Our data address the first part of that question” and “We did not test that possibility.” These are substantive answers when they accurately describe the study. For a Japanese practice guide, see 学会での英語質疑応答.

Collaboration includes the apparently ordinary language

Projects also need clear messages about deadlines, responsibilities, revisions and disagreement. A brief written confirmation can prevent an unresolved assumption from becoming a disagreement several weeks later.

Practise an email that names the proposed action, its owner, the deadline and anything still undecided. For example: “I can revise the methods section by Friday. Could you confirm which dataset we are reporting? I will leave the results unchanged until we agree.” This is an original practice example. Its value lies in the decisions it makes visible.

The evidence for unequal burdens needs its own boundaries

Ramírez-Castañeda’s 2020 study surveyed 49 Colombian researchers in biological sciences who had completed or were undertaking doctoral study. Participants reported additional burdens when preparing work in English, including writing time and financial costs. The time estimates were self-reported. This is evidence from a specific group, not a measured surcharge for every researcher worldwide.

A different problem appears in Amano and colleagues’ 2016 study. Their searches in 16 languages identified 75,513 documents on biodiversity conservation published in 2014; 35.6% were in languages other than English. That figure concerns this search and subject area. It is not the proportion of all science published outside English. It does show why assuming that all relevant evidence will be discoverable in English is unsafe.

These studies support different concerns: the burden of participating in English and the loss of knowledge when searches exclude other languages. Teaching researchers English addresses part of the first concern. It cannot, by itself, solve the second.

Make support part of the research process

I would build support around a real document or interaction: an abstract, a methods explanation, a conference question or a message to collaborators. Agree on the meaning, revise the language, then check what another reader or listener understands.

Institutions and research teams can contribute through editing support, time for rehearsal, multilingual summaries and explicit plans for relevant non-English literature. UNESCO’s 2021 Recommendation on Open Science includes multilingual knowledge within its account of open science. This is a policy principle, not evidence that every institution already provides such support.

The aim is research that another person can assess accurately, with its contribution and limitations intact. That is a more useful standard than making every researcher sound as though they grew up in the same place.

Sources and further reading

1. Ramírez-Castañeda, V. (2020). Study of English-publication burdens among Colombian biological-science researchers. PLOS ONE, 15(9), e0238372.
2. Amano, T., González-Varo, J. P., & Sutherland, W. J. (2016). Languages Are Still a Major Barrier to Global Science. PLOS Biology, 14(12), e2000933.
3. UNESCO (2021). Recommendation on Open Science.

Teaching examples and recommendations are the author’s proposed practice, not interventions evaluated in these studies. Prepared with AI-assisted editing. More writing by Mark Zolotar.