Checking the Web for Public Facts: Can We Do This Together?

• In Event: Citizen Bureaucraft - I

Abstract

Fact-checking has the potential to organise collective intelligence through the shared evaluation of public claims (Graves, 2016). Such efforts may be able to reinvigorate knowledge culture, but existing methods suffer from limitations: they are not well equipped to address emergent statements in fast-moving and contested areas of public knowledge formation. Here, debunking statements may aggravate polarizations and be perceived as patronizing and excluding citizens (Marres, 2018). This paper will narrate a methodological experiment conducted in collaboration with the Public Data Lab which aimed to prototype new methods for discovering, accounting for and responding to a specific type of public knowledge claim which we call "experimental facts": claims in the process of being formulated, with an unstable truth value. Working with social researchers, designers, fact-checkers, journalists and engineers, we would like to develop test cases for experimental fact-checking and use these to test and refine a combination of moderated, crowdsourced and automated approaches to fact-checking, as a contribution to collective learning and the strengthening public knowledge culture.

Author

• Noortje Marres, University of Warwick, UK