Social media allow democratization and immediate access to information. Despite this, a significant disadvantage of these platforms is the dissemination of fake news. Fake news is usually presented in the form of text; therefore, it is relevant to investigate it with linguistic, content-based analysis, to better understand how and why users decide to share the content regardless of its trustworthiness. To this date, few studies have a qualitative approach and most linguistic analyses of fake news often use machine learning or deep learning techniques. To overcome this shortcoming and improve fake news recognition, this research aims at analysing 50 racial hoaxes selected from debunking websites in Italy (Bufale.net, Butac. it), with a quanti-qualitative approach. Hoaxes’ titles were collected, classified, organized, and analysed using psycholinguistic methodologies. The study aims to identify the language features that make racial hoaxes engaging and explore their variation across different types of threats. The main results of the study pointed out how racial hoaxes are characterised by the widespread use of provocative emotional content; moreover, more negatively polarised words, more vulgar expressions and in general, a more aggressive connotated language were found in criminality threat hoaxes compared to the others.
THE LANGUAGE OF RACIST VIRALITY: A CONTENT ANALYSIS OF ITALIAN RACIAL HOAXES
Scardigno R.;D'errico F.
2024-01-01
Abstract
Social media allow democratization and immediate access to information. Despite this, a significant disadvantage of these platforms is the dissemination of fake news. Fake news is usually presented in the form of text; therefore, it is relevant to investigate it with linguistic, content-based analysis, to better understand how and why users decide to share the content regardless of its trustworthiness. To this date, few studies have a qualitative approach and most linguistic analyses of fake news often use machine learning or deep learning techniques. To overcome this shortcoming and improve fake news recognition, this research aims at analysing 50 racial hoaxes selected from debunking websites in Italy (Bufale.net, Butac. it), with a quanti-qualitative approach. Hoaxes’ titles were collected, classified, organized, and analysed using psycholinguistic methodologies. The study aims to identify the language features that make racial hoaxes engaging and explore their variation across different types of threats. The main results of the study pointed out how racial hoaxes are characterised by the widespread use of provocative emotional content; moreover, more negatively polarised words, more vulgar expressions and in general, a more aggressive connotated language were found in criminality threat hoaxes compared to the others.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
