Detecting Linguistic Characteristics of Political Disinformation in Indonesian Social Media: Insights From Systemic Functional Linguistics

Abstract

This study examines the linguistic features of political disinformation on Indonesian social media during the 2024 presidential election using Systemic Functional Linguistics (SFL). The study employed a qualitative descriptive method, collecting data from Twitter (now X) and Instagram posts related to the 2024 Indonesian presidential election, focusing on posts containing clear evidence of disinformation. The analysis mapped the lexicogrammatical features of disinformation at three levels: ideational, interpersonal, and textual metafunctions to explain how these features shape the overall discourse. The study reveals frequent use of Material and Relational processes to misrepresent political figures, while Verbal processes distort statements through selective quoting. Attitude, engagement, and graduation features are also prominent, with disinformation posts expressing strong negative judgments about political opponents. Engagement techniques, such as selective citation and heteroglossia, create an illusion of balanced argument, while graduation features amplify emotional intensity through exaggerated language and forceful assertions. Disinformation posts rely on declarative clauses, rhetorical questions, and high modality to present falsehoods as factual, while causal conjunctions and marked themes enhance the coherence of biased narratives. The study underscores the need for a metalinguistic approach to social media literacy, equipping users with tools to critically analyze disinformation.