Moralized language predicts hate speech on social media

Abstract

Hate speech on social media threatens the mental health of its victims and poses severe safety risks to modern societies. Yet, the mechanisms underlying its proliferation, though critical, have remained largely unresolved. In this work, we hypothesize that moralized language predicts the proliferation of hate speech on social media. To test this hypothesis, we collected three datasets consisting of N = 691,234 social media posts and ∼35.5 million corresponding replies from Twitter that have been authored by societal leaders across three domains (politics, news media, and activism). Subsequently, we used textual analysis and machine learning to analyze whether moralized language carried in source tweets is linked to differences in the prevalence of hate speech in the corresponding replies. Across all three datasets, we consistently observed that higher frequencies of moral and moral-emotional words predict a higher likelihood of receiving hate speech. On average, each additional moral word was associated with between 10.66% and 16.48% higher odds of receiving hate speech. Likewise, each additional moral-emotional word increased the odds of receiving hate speech by between 9.35% and 20.63%. Furthermore, moralized language was a robust out-of-sample predictor of hate speech. These results shed new light on the antecedents of hate speech and may help to inform measures to curb its spread on social media.

Believability and Harmfulness Shape the Virality of Misleading Social Media Posts

“…Limitations and future work: As with others, our study is not free of limitations and offers potential for future work. First, analogous to earlier observational studies [e. g., 41,47,48,54], we demonstrate associations and not causal paths. Second, experimental studies in controlled settings may help to understand whether the perceptions regarding the believability and harmfulness of misinformation differ between community fact-checkers, experts, and regular social media users.…”

Bases sociocognitivas do discurso de ódio online no Brasil: uma revisão narrativa interdisciplinar

“…Alinhando-se ao que propôs Silva (2018) sobre a violência, aspectos morais e emocionais têm sido, assim, apontados como preditores de engajamento online Wills et al, 2017), porque os traços morais, mais do que as próprias experiências pessoais (memória autobiográfica), são apontados, na literatura, como a parte mais essencial da identidade de uma pessoa, aquilo que a configura como um "eu" (Strohminger;Nichols, 2014). O ódio, como uma emoção moral (Smith, 2015), cumpre papel fundamental no engajamento das pessoas em publicações online (Solovev;Pröllochs, 2023), carecendo, por isso, de atenção especial, uma vez que as redes sociais mobilizam comportamentos agressivos tanto dentro quanto fora do espaço virtual, sob o risco de transformar tais comportamentos em um padrão socialmente aceito.…”

“…Compreender esse fenômeno social requer esforço interdisciplinar, porque, particularmente na Internet, esse comportamento proporciona riscos graves à segurança e à saúde mental dos cidadãos em diversas sociedades modernas (Solovev;Pröllochs, 2023), ao ponto de estudos sobre o ódio migrarem para o domínio da saúde pública, em face de sua inerente relação com a violência (Krug et al, 2002). Combater discursos de ódio é urgente para proteger direitos humanos, democracias, inclusão e respeito à diversidade, com especial atenção aos movimentos radicais e extremistas violentos, que têm, entre outros fatores, o ódio em suas raízes (Sternberg, 2005).…”