As a researcher, there are few things more vital to your reputation than your publishing record. But what if your native language is not English? Is this a significant problem? How can it be mitigated? Nishchay Shah, CTO of CACTUS, investigates.

According to data compiled by Scott Montgomery in Does Science Need a Global Language, more than three-quarters of scientific papers are published in English, and in some fields, this figure exceeds 90%.

Whether your objective is tenure, a research grant, a book deal, or just an improved reputation in your department, a string of high-profile journal articles are the currency you require to trade your way to the top. Because the global publishing language is English, this success depends on being able to articulate yourself well in what may not be your first language.

Considering the need to publish in English, not being a native English speaker could be holding 95% of the world’s population back. It’s like a language tax levied by the native English-speaking world on the non-native population, one that is paid with increased effort and a higher rate of rejection.

Impact of the language tax

In a 2018 study published in Written Communication, Hanauer et al. investigated the added burden faced by Mexican and Taiwanese researchers when writing research articles in English. They found that non-native English speakers had an average 24% increase in difficulty, 10% rise in dissatisfaction and 22% more anxiety when it came to manuscript preparation. The researchers maintain that the additional burden of English as a second language constitutes a linguistic injustice — a barrier to science that needs to be addressed.

The primary problem for non-native English speakers is that writing a research paper in English takes them longer. On the other hand, the surge in research output means they need to do more to keep up with developments in their field. An article published in Nature points to bibliometrics showing an annual 8-9% increase in published scientific papers in recent decades. In fact, over a million biomedical research papers are added to PubMed every year, which is approximately two papers per minute. Keeping up with such an overwhelming volume of research is challenging for all researchers, but even more so for those who must read and understand this in English, their non-native language.

Once non-native English speakers begin writing, grammatical woes can slow down the process even further ─ English grammar is notoriously fickle. While even native English-speaking researchers may struggle with parallelism or infinitives, navigating these complex grammar rules in a second language is much more difficult. In competitive research fields, taking longer to publish may mean another researcher beating you to the punch.

Desk rejection and the language tax

The language tax increases the risk of desk rejection for three key reasons. The first, and most obvious, is that if the author’s writing and grammar skills are not as strong as their academic expertise, the manuscript could be difficult to read or understand. The second reason for rejection is that a non-native English speaker might find it more difficult to express their ideas in English, resulting in a manuscript that does not showcase or convey their research effectively. The third reason for desk rejection is not complying with the rules and regulations required to submit successfully, such as the journal’s style guide and referencing format; these guidelines are usually produced in English, making it more difficult to understand and follow.

However, in a 2013 study published in BioScience called Predicting Publication Success for Biologists, William F. Laurance et al. suggested that native English speakers, those who published earlier in their careers, and males have only ‘minor advantages’ in the long term. The biggest benefit identified by the paper is a good pre-PhD publication record, and its findings are sound in that regard. It’s important to note here that this paper looked at long-term career success in relation to a researcher’s publishing record and not their likelihood of getting into print in the first instance.

Seeing success on your first try is still far more common if you are a native English speaker. So, while getting published early is seen as the most important indicator of success, achieving this requires a high degree of familiarity and quick mastery of the English language. To my knowledge, research is still to be done on whether an author is more likely to be published at pre-PhD stage if their native language is English.

Overcoming the language tax

Editing tools and services offer a way to bridge the language gap for non-native English speaking researchers by getting their writing up to the standard a journal requires. However, while there are many tools available to check spelling and grammar online, they are usually not designed to support scientific content and, therefore, are not the best choice to improve academic writing. Moreover, there are few tools that help the author as they work on the all-important first draft and even fewer that conduct the kind of technical checks required to ensure a manuscript complies with journal-specific submission guidelines. Getting their manuscript written, polished and through to the final stage of the process is something many researchers struggle with.

Specialised assistive writing tools, based on artificial intelligence (AI) and machine learning, emerge as the natural next step. For example, Paperpal’s AI-powered assistive writing technology can offer language suggestions while a researcher writes, offering feedback and tips to improve grammar, punctuation, style, and readability. With real-time suggestions to polish academic writing, the tool can help speed up the overall writing process and reduce the need for extensive revision, which instils a sense of confidence in the researcher.

It also allows journals to support researchers with English as a second language by giving them the opportunity to check their manuscripts against journal-specific requirements, such as key declarations, style and structural issues, or inconsistencies in references, figures and tables. By hosting this AI tool on their website for researchers to use during the submission process, journals can also ensure they don’t miss out on significant research that may be rejected due to often avoidable oversights.

While the increasing popularity and success of such software has raised the question of AI potentially replacing writers completely, most academics feel that the creativity, experience and technical expertise humans can work into their writing cannot be replicated by AI.

Artificial intelligence is a transformative technology for academic writing — as important for academics attempting to achieve publication as the moment when the typewriter was replaced with the word processor or snail mail with email. I hope it will help the academic community put an end to the language tax for good, democratising scientific publishing and making English language journals significantly more accessible to non-native speakers.