The Gate That Moved: arXiv's Endorsement System, Algorithmic Popularity Bias, and the Conservation of Gatekeeping in Open Science

This volume turns the Consolidation Trap on the institutions this research program's author has personally relied on to circulate its own findings. The volume's central formal contribution is the Conservation of Gatekeeping: in any system receiving submissions at a volume exceeding what unmediated attention can process, some filtering mechanism must emerge to allocate the genuinely scarce resource — reader attention, which does not scale — and removing one filtering mechanism without resolving the underlying volume-to-attention ratio does not eliminate gatekeeping but displaces it, typically with less visibility and less contestability. The volume grounds this in arXiv's thirty-five-year history: founded in 1991 as an explicit escape from informal, personally-mediated preprint circulation, arXiv introduced its endorsement system in January 2004, a mechanism structurally continuous with guild apprenticeship and the Royal Society's gentlemanly witness, compressed into thirteen years. Most currently, in December 2025 and January 2026, arXiv removed institutional email affiliation as a sufficient basis for automatic endorsement, citing an unsustainable volume of non-scientific submissions, drawing a public comment naming the resulting system private gatekeeping. The volume's second empirical pillar is the popularity-bias literature in recommender systems and academic search ranking, mathematically continuous with the preferential attachment used to formalize faculty hiring concentration, and documented explicitly in academic paper ranking through venue-popularity weighting (Sun & Giles, 2007). A dedicated section addresses this program's own question, raised in its preface, of whether its central claims had already been reached elsewhere: it engages Hugh Desmond's (2024) independently argued case that gatekeeping should be conserved in the open-science era, and the broader diagnosis, from the general gatekeeping literature (Barzilai-Nahon 2009) and from a 2026 landscape survey of metascience (Watney & Gustetic 2026), that no unifying account of these mechanisms yet exists. The volume closes by subjecting the author's own primary circulation channel, PhilPapers and PhilArchive, to the same scrutiny, concluding that editorial screening constitutes a different gate rather than an absent one.