Astronomy Preprints on arXiv Quadruple Since 2015, Bypassing Peer Review and Burdening Early-Career Researchers

The preprint quantifies how arXiv's original 1990s architecture, built for low-volume circulation, now processes thousands of monthly astro-ph uploads without version control or review flags. Researchers must manually filter multiple iterations of the same study, a load that falls heaviest on postdocs and graduate students who lack journal access or collaboration buffers. This matches patterns seen in high-energy physics, where unvetted preprints accelerated claims later corrected in journals.

Bedin Rolly proposes separating Level-1 peer-reviewed papers from Level-2 drafts to restore precedence norms and let users opt out of draft streams. The suggestion addresses a gap missed in earlier coverage: the incentive distortion that pushes junior scientists into large consortia rather than independent work. Similar bifurcations already operate in parts of computer science via conference tracks.

Data from the decade-long metadata scrape reveal the cognitive tax has grown faster than submission volume alone would predict because authors increasingly post iterative versions. Early-career impacts include reduced time for novel projects and heightened risk of being scooped by unverified claims. Strengthening evidence would require linking individual researcher productivity metrics to preprint posting frequency.

Next steps hinge on arXiv governance adopting metadata tags for review status, a change that could be piloted in astronomy before platform-wide rollout.