Fri Mar 22

Prestigious journals make it hard for scientists who don’t speak English to get published. And we all lose out

Written by: Henry Arenas-Castro, Postdoctoral Fellow, Yale University, The University of Queensland

For the first time in history, a single language dominates global scientific communication. But the actual production of knowledge continues to be a multilingual enterprise.

The use of English as the norm poses challenges for scholars from regions where English is not widely spoken. They must decide whether to publish in English for global visibility, or publish in their native language to make their work accessible to local communities. And when they work in English, they end up expending more time and effort ^[1] writing and revising papers than their native English-speaking peers.

As gatekeepers of scientific knowledge, academic publishers play a key role in helping or hindering the participation of a multilingual scientific community. So how are they doing?

We reviewed the policies of 736 journals in the biological sciences and discovered the great majority are making only minimal efforts to overcome language barriers in academic publishing. Our research is published in Proceedings of the Royal Society B ^[2].

A wide range of inclusive policies

Linguistically inclusive policies come in many forms, and can be implemented at each stage of the editorial process. They might aim to make publishing more multilingual. Alternatively – if sticking with English – they may aim to reduce the burden on non-native English speakers.

Allowing papers to be published in more than one language at the same time would resolve the dilemma many non-native English speaking scholars face about communicating locally or globally. However, only 7% of the journals we surveyed allowed this possibility. (A further 11% will allow multilingual versions of an abstract alone.)

Another possibility would be to implement machine translation tools to make versions of an article available in multiple languages on a journal’s website. There has been recent progress in this area ^[4], but only 11% of journals we surveyed have put it into practice.

Journals can also indicate they value submissions from authors from diverse linguistic backgrounds by explicitly declaring they will not reject manuscripts solely on the basis of the perceived quality of the English. Surprisingly, we found only two journals stated this.

Similarly, providing author guidelines in multiple languages would further encourage submissions from diverse authors. While 11% of the journals we examined translate specific sections of their guidelines to other languages, only 8% offer their entire guidelines in more than one language.

To ensure published research learns from the scientific contributions of scholars from around the globe ^[5], journals should explicitly allow or encourage non-English literature to be cited. Only one tenth of journals mention this in author guidelines.

Journals may also adopt measures to ensure work submitted by non-native English speakers is assessed fairly. One such measure is the provision of English-language editing services.

More than half the journals we surveyed refer authors to some kind of editing services; only 1% offer the service free of charge to authors. The cost of editing may impose a considerable financial burden ^[6] on scholars in lower-income countries.

Another measure is to educate reviewers and editors about language barriers and instruct them to assess the manuscripts based on their research attributes alone. This is something only 4–6% of journals implement.

Drivers of inclusivity

We also identify two key influences on a journal’s adoption of linguistically inclusive policy.

The first is impact factor, a measure commonly taken to represent the prestige of a journal. We found journals with higher impact factors tend to adopt less-inclusive policies, possibly because they mostly target English-proficient authors and readers.

The second influence is ownership by a scientific society. Journals owned by scientific societies tended to adopt more inclusive policies. They have also taken the lead in the movement to publish multilingual content.

Many scientific societies have a mandate to foster diverse communities ^[8]. They are supported by their members and are well positioned to push for a cultural change in scientific publishing.

We also found that open access journals (which make research available to the public for free) were no more likely to adopt inclusive linguistic policies, nor were journals with more diverse editorial boards.

The apparent lack of influence of linguistically diverse board members is a puzzle. Perhaps editors who have experienced language barriers in their own professional life do not advocate for non-native English speaking authors. Or perhaps editorial boards have less power to define editorial policies than we might expect.

Language barriers

Language barriers deepen geographic divides, hampering knowledge sharing. Tackling them in academic publishing becomes critical to effectively address both regional and global issues, such as health and conservation.

In our study, we looked at a number of linguistically inclusive policies, but there are plenty of other things journals can do to help scientists from non-English speaking backgrounds. These range from using artificial intelligence tools ^[9] to the re-negotiation of copyrights to authorise the publication of translations ^[10] elsewhere.