Preprints and the Version of Record: A Reader Problem

Abstract

Where several versions of a paper circulate, readers must decide which one to trust and cite. We survey versioning practice across preprint servers and journals, and argue that the burden currently placed on readers belongs with infrastructure.