Rejected for the English

A paper can be turned down for its science. It can also be turned down for its English. The second is a different kind of rejection, and it does not fall evenly.

Amano et al., writing in PLOS Biology in 2023, surveyed 908 environmental scientists in eight countries: Bangladesh, Bolivia, Britain, Japan, Nepal, Nigeria, Spain, and Ukraine. Among early-career researchers, 38.1% of those from countries of moderate English proficiency had a paper rejected because of the English. For native speakers, the figure was 14.4%. That is about 2.6 times as often.

The cost does not stop at the journal. Researchers from countries of moderate English proficiency spend a median 93.7% more time preparing an oral presentation in English. Among early-career non-native speakers from high-income countries, about 30% avoid English-language conferences because of the language, and about half avoid giving talks.

None of this measures the quality of the work. It measures where the work was written.

The same paper suggests that journals and universities allow the appropriate use of AI tools to reduce these barriers. The word that matters there is appropriate. A tool that changes what the author meant is not a fix. It is a second problem.

Tolayda is built around that line. The English sits beside the original, so the author can check every passage against what they wrote. Export waits until the author trusts it. Scholarly review, copy-editing, and any official certification stay with the people responsible for them.

The writing-time figures from the same study are in the note on the tax.