In 100 minutes I'm going to be interviewed by Physics Today about the arXiv's new requirement that every paper either be in English or include an English translation (https://blog.arxiv.org/2025/11/21/upcoming-policy-change-to-non-english-language-paper-submissions/). What should I say?
Here's what I plan to say:
1) This policy would have ruled out important works like Grothendieck's SGA1 (https://arxiv.org/abs/math/0206203). It's unlikely that someone would translate a 327-page book just for the privilege of putting it on the arXiv.
2) Some scientists say now they'll switch to using HAL (https://hal-hprints.archives-ouvertes.fr/), a European open-access repository that accepts articles in every language.
3) This move seems to hint at a growing "Americanization" of the arXiv, along with their September 2024 switch to no longer supporting mirror websites in countries outside the US (https://blog.arxiv.org/2024/09/13/attention-arxiv-users-arxiv-mirrors-to-shut-down-september-15th-2024/). It seems the arXiv is planning to move their storage from Cornell University to a "Google cloud platform" (https://telescoper.blog/2025/04/24/arxiv-the-cloud-and-backups/). Europeans responded by creating a "dark archive" for the arXiv: a backup that's not publicly accessible but can be brought online if the forces of evil crush the US arXiv (https://telescoper.blog/2025/05/14/an-archive-for-arxiv/).
4) It's a separate issue, but Physics Today should be aware that the arXiv is getting lots of AI slop papers. Maybe a topic for another article!
