This paper provides a closer analysis of the discourse in doxing in a sample of digital and printed US and European media with a particular focus on the Capitol riots. The analysis centers around the following questions: How is doxing portrayed? How are its victims and perpetrators portrayed? What expressions about doxing appear depending on who performs the act versus being exposed? When it comes to how doxing is valued in the discourse, there are three categories that determine how the discourse is portrayed. (1) Who is behind the doxing? (2) who is the target of doxing? (3) What is the purpose of doxing? These categories work in symbiosis with each other. A positive notion of the doxxer, a negative notion about the person being doxxed and a justified purpose is needed for it to be valued in the discourse. Two main types of doxing could be distinguished that are framed in different ways in the discourse, doxing for malicious purposes and doxing for political purposes. In relation to the U.S capitol riots, Doxing was recontextualized. The change is explained trough (Re-) definition, a ‘theoretical legitimation’ strategy where actions are legitimized through defining an action in terms of ‘another, moralized activity’.

Google Translate’s mission is “to enable everyone, everywhere, to understand the world and express themselves across languages” (Pitman, 2021). But are algorithms capable of leading us beyond the translation of the word toward an understanding of the world? Computational linguistics research has been interested in assessing this kind of real-world effects of technology and invited other disciplines to join their effort. With this purpose, this study examines the ways the algorithmic agency (Maly, 2022) elicits a ‘movement of meanings’ (Silverstone, 1999) when mediating news reports from English to Portuguese – the official language of Brazil, the country with the greatest use of Google Translate (Turovsky, 2016). For that, it investigates how algorithms convert appraisal and semiotic elements that carry ideological stances. The bilingual sample consists of six news articles on the U.S. Capitol attack published in U.S. outlets, two each of right, center, and left political leaning, along with their translations obtained through Google Translate. The analytical framework encompasses Fairclough’s (2003) CDA methods that allow an exploration of how discourses embedded in these texts represent the social phenomena that are being depicted. This lens is complemented by the Appraisal theory (Martin & White, 2005) to investigate how value positions are constructed within texts through evaluation. A third analytical tool is necessary to engage with the ways in which meanings are moved from source to target texts. For this, van Leeuwen’s (2008) notion of recontextualization affords an assessment of the processes inherent to translations. The analysis showed that algorithms neutralized appraisal through lexical choices, changed semiotic elements through recontextualization, and blurred stances by standardizing the target language. The paper, thus, concludes that Google Translate constructed power by renaming reality and enacted it by reshaping evaluations, advancing research that seeks to examine algorithms’ impacts on digital discourse. Speaking from the epistemic locus of the Global South, this thesis proposes a critical reflection on the ideologies concealed by the self-proclaimed discourse of the universality of digital technologies.

No description available.