The linguistic structure of fake news in advertising posts of social media

Investigating fake news has a long history, almost as old as the history of media. This issue has recently returned to the attention of researchers. Since the raw material of all types of news is linguistic content, the characteristics of fake news from the linguistic point of view have recently been the focus of linguists. In this research, which is descriptive-analytical in terms of methodology and library-type in terms of data collection, based on the framework proposed by Grieve and Woodfield (2023), we investigate the linguistic features of fake news, in a corpus of 100 advertising text messages. Data have been selected from the Persian language pages of Telegram and Instagram social messengers in the fall and winter of 1402. The purpose of this research is to investigate the linguistic components of fake news in social messengers. The analyzed corpus includes messages that are specifically classified as cases of disinformation against misinformation. The 7330- word corpus of this research, includes messages that are apparently intended to inform and advertise, but actually to deceive the audience on popular social media pages and channels. The results showed that these types of messages are instances of low-density messages in terms of information density, but high-conviction texts in terms of conviction. The findings of this research can lead to the deepening of linguists' insight into the linguistic mechanisms of fake news (especially misinformation) and ultimately lead to an increase in the awareness of the audience of such media and messengers.