arXiv’s new preprint submission cap divides opinion

The preprint server arXiv has introduced a cap of two submissions per author per month in response to an “exponential increase” in volume, drawing both praise and criticism from scholars.

In a blog post published on 1 October, arXiv, which primarily publishes papers in physics and mathematical subjects, explained that the change comes as the platform, and the wider scientific community, are “facing a watershed moment” amid an AI-enabled surge in research papers.

The open access repository has seen its submissions double in the past two years alone, receiving 40,363 submissions last month, compared with 9,869 submissions in September 2016.

Although, as a preprint server, arXiv does not carry out traditional peer review, which evaluates the scientific quality of submissions, all submissions go through a “moderation” process, which “strives to rapidly disseminate research while meeting content and technical standards”. A small team of staff and students “work with 170+ volunteer moderators to review the classification, metadata, and content of all submissions”, its submission guidance explains.

“The team performs checks with software support to confirm that the submission is classified to the appropriate subject category, that it is in compliance with our technical requirements, including formatting and metadata information, and that the content is appropriate, which, generally, describes content that would be refereeable by a conventional journal and in the format of a research article.”

The strain on the volunteers who carry out this moderation means that the preprint server is now limiting individual researchers to “up to two submissions per calendar month, with a limit of three total active submissions at any given time”, the blogpost explains. “This policy update is being put into effect to equitably distribute our volunteer moderators’ time across arXiv authors and protect the arXiv corpus from a sharp increase in inappropriate submissions.”

arXiv editor-in-chief Thomas Dietterich said on X: “Analysis of submissions shows that some authors are submitting dozens of papers. This takes moderation resources away from the rest of us, which causes delays in our papers being released.”

On social media, scholars expressed mixed feelings about the announcement.

One opponent was Scott Kominers, Sarofim-Rock professor of business administration at Harvard Business School. He said on X that arXiv would do better to “economise on moderation – not on scholarship. As a market-design scholar and arXiv contributor of almost 20 years, I think this policy confuses the two.”

He conceded that arXiv moderators “face a genuinely hard problem, and I don’t question their commitment. But taking their resource constraints seriously is precisely why we need a better mechanism for allocating their attention.”

Kominers, whose volume of submissions has previously been restricted by the platform, conceded that quotas have “a plausible benefit” in that they encourage authors to be selective.

“But selecting papers likely to pass moderation isn’t the same as selecting for scientific value. And deciding which papers to submit is only part of the problem,” he said. Instead, the “crucial question is which submissions require attention – and when review is worth its cost”.

However, many scholars have welcomed the new policy. One of them is Alex Kontorovich, distinguished professor of mathematics at Rutgers University. On X, he said: “I actually like this policy. I’m not going to read your 42 papers this month. Pick two, your very best two.”

Echoing his sentiment, Catherine Hemans, Astronomer Royal for Scotland, said on Bluesky: “I’m so confused. Who would want to submit so much slop? For me – [in] both writing and hiring – it’s always been about quality over quantity, every single time.”

But Dare Obasanjo, a software engineer, lamented the need for arXiv’s decision, saying on Bluesky that the move signalled “yet another open submission process killed by LLMs (large language models)”.

frances.jones@timeshighereducation.com