Who Decides What is 'True'? Aos Fatos, Zuckerberg, and the Paradoxes of Fighting Disinformation

Abstract

This article examines two institutional documents from the fact-checking organization “Aos Fato”’ website, aiming to analyze (i) how language performs the agency’s authority, and (ii) which social voices are privileged or authorized in its discourses. The theoretical and methodological foundation of the article is based on Sociodiscursive Interactionism (SDI) (Bronckart, 2006); the New Literacy Studies (NLS) (Street & Castanheira, 2014) and Media Literacy (Hobbs & Jensen, 2009; Ferreira, 2021); studies on disinformation (Komesu, Alexandre, & Silva, 2020; Assis, Komesu, & Pollet, 2021; Wardle & Derakhshan, 2023); and the concept of truth as discussed by Foucault (1996) and Arendt (2011). From Foucault (1996), the concept of “will to truth” is taken as an exercise of discursive power that guides the analysis of how the agency’s verification criteria (metrics, classificatory labels, and “reliable” sources) construct an authority that naturalizes certain truths in the digital public sphere. From Arendt (2011), the thesis on the fragility of factual truth is incorporated, which, due to its reliance on testimony and records, is reduced to mere opinion in political contexts. The article uses the textual analysis model proposed by Bronckart (2006), which articulates linguistic and discursive dimensions, to verify how the agency’s language performs authority, which social voices are privileged, and under which social and historical conditions this construction is legitimized. The results provide clues that, although fact-checking organizations like “Aos Fatos” represent a significant advancement in combating disinformation, their editorial policy reveals some vulnerabilities regarding verification criteria and the authority discursively constructed by the institution, which seems to point to the need for other strategies capable of addressing the roots of the informational crisis threatening contemporary democracies.