ABSTRACT
The article focuses on hostile language used in electronic comments, published in a news site, after the Brazilian electoral campaign of 2014. To this end, the concepts of flaming (Kayany 1998) and polemical discourse (Amossy 2011) are introduced. For corpus analysis, Bousfield´s (2008) categories for the study of impoliteness in language are adopted. Negative and positive face threats were identified in the corpus, within the scope of polemical discourse. The research findings suggest that interactional categories stand a better chance towards the characterization of the phenomenon of flaming than the exclusive use of lexical categories.
Key-words:
flaming; impoliteness in language; polemical discourse