Continuity and discontinuity in web archives: a multi-level reconstruction of the firsttuesday community through persistences, continuity spaces and web cernes

Abstract

Memory institutions have been archiving the Web for the last 25 years. These initiatives seek to preserve pieces of our digital heritage by harvesting web resources. But just like Funes in Borges short story, memory institutions will never be able to achieve exhaustiveness. Archiving always goes with complex selection criteria stated by librarians, curators, politicians or – in the particular case of web archiving – engineers and robots called crawlers that determine the spatio-temporal coverage of archived corpora. Unlike traditional archival materials, web archives can’t be understood apart from their own archiving processes : crawlers tear web resources away from the continuous temporality of the Web and produce discretized snapshots timestamped by archiving date. By nature, Web archives are not direct traces of the Web, they are direct traces of crawlers (1). The web archives movement has originally been sparked with the intuition that web resources were intended to become valuable research materials in the hands of future historians ; and archiving pioneers indisputably succeeded