Analysis of Semantic and Syntactic Features of Cyberbullying on Instagram and X: Iranian Users

Cyberbullying is a widespread and harmful issue, especially with the rapid growth of social media in Iran. It causes significant psychological distress for victims, making it a critical area for research. This study analyzes the linguistic, semantic, and syntactic features of cyberbullying content produced by Iranian users on Instagram and X (formerly Twitter). Focus areas include insulting vocabulary, derogatory metaphors, dominant themes, and syntactic strategies, to understand how language is used to weaponize aggression in the Persian context. A purposive sample of 500 posts and comments (250 per platform) was collected over a specific period, selecting content with clear or implied indicators of cyberbullying, such as hostile language or threats. Qualitative content analysis with thematic coding was conducted using MAXQDA, complemented by corpus linguistics techniques to examine key lexical patterns, metaphors, and punctuation. Particular attention was given to passive constructions, agent omission, and punctuation (e.g., repeated exclamation points) as strategies to evade accountability. Results revealed three major trends: Gendered insults comprised about 45% of offensive language, often targeting women with sexualized slurs or stereotypes. Emphatic punctuation, especially multiple exclamation marks (!!!), was used in 65% of cases to intensify hostility. Around 70% of content employed ambiguous syntactic structures—such as passive voice or missing subjects—to conceal the bully’s identity and reduce responsibility. Additionally, cultural metaphors comparing victims to animals or objects dehumanized them further. The study emphasizes that Persian cyberbullying relies heavily on syntactic ambiguity and gendered slurs, which pose challenges for detection tools typically designed for English. Developing localized AI-based algorithms trained on Persian data, capable of recognizing these specific markers, is essential. Policies and social media platforms should adopt context-aware solutions and educate users to combat this rising issue. Future research should include regional dialects and visual content analysis to address cyberbullying comprehensively in Iran.