Automating the Horae: Boundary-work in the age of computers

This article describes the intense software filtering that has allowed the arXiv e-print repository to sort and process large numbers of submissions with minimal human intervention, making it one of the most important and influential cases of open access repositories to date. This article narrates arXiv's transformation, using sophisticated sorting/filtering algorithms to decrease human workload, from a small mailing list used by a few hundred researchers to a site that processes thousands of papers per month. However, there are significant negative consequences for authors who have been filtered out of arXiv's main categories. There is thus a continued need to check and balance arXiv's boundaries, based on the essential tension between stability and innovation.