Textos digitais e manipulação – a construção dos sentidos em narrativas desinformativas

Abstract

This thesis investigates how textual criteria can help explain manipulation in disinformative narratives shared in digital environments, based on the theoretical and methodological approach of Textual Linguistics. To support the study, we use the following theoretical perspectives: the concept of text and reference by Cavalcante et al. (2019, 2020, 2022); the referential networks proposed by Matos (2018); the enunciative-interactional framework of Textual Linguistics by Cavalcante, Brito, and Martins (2024); the manipulation strategies described by Charaudeau (2020); the the conception and characteristics of disinformation by Recuero (2024); and the post-dualist view of language and technolinguistic gestures by Paveau (2021). These ideas are combined in a focused interdisciplinary approach (Charaudeau, 2013) to understand how manipulation strategies are built through language in disinformative narratives in digital media.This is an exploratory and qualitative study that analyzes six cases of disinformation (including both static and dynamic texts) published in different digital platforms, such as TikTok reels, Instagram posts and posts on X. All of them deal with the Brazilian political context. The analysis follows these steps: contextualization and fact-checking based on the work of agencies like Lupa, Aos Fatos, Projeto Comprova, UOL Confere, and Estadão Verifica; development of the theoretical-methodological framework to identify patterns in disinformative texts; analysis of discourse objects through referential networks as evidence of manipulation strategies; and observation of meaning effects created by technolinguistic gestures, such as hashtags, emojis, comments, and interactive buttons. Results show that the speaker often plays the role of an impostor, changing meanings in the interaction by editing or cutting news content and combining it with different multimodal resources (like audio and captions that represent the identity of certain groups) and technolinguistic gestures (emojis, hashtags, comments). The (re)construction of referents in different connections reveals manipulation strategies based on dramatization (presenting social disorder, a guilty party, and a savior) and on the promotion of values (family, country, work, victimization of the people, demonization of enemies, and populist leadership). These strategies create effects such as affective polarization, conspiracy theories, mistrust in institutions, political binarism, herd behavior and identity belonging which support the manipulation of public opinion, which science must vehemently combat.