Digital thorns: the rise of sarcasm as political protest in Türkiye dijital dikenler

Abstract

Sarcasm, by revealing unforeseen linguistic and societal interactions in online platforms dominated by political debates, presents a critical area of study for natural language processing models. The intensification of sarcasm in digital discourse has become a subject of interest, as increasing censorship and authoritarian tendencies have amplified the need for encoding dissident thought. The most significant finding of this study is that the high consistency rates achieved by a BERTurk model on the Ekşi Sözlük dataset demonstrate sarcasm’s potential as a real-time indicator of societal discontent and political tension. The model revealed a stark asymmetry in sarcastic content: entries targeting Recep Tayyip Erdoğan (ruling party leader) exhibited a higher sarcasm rate (28.1%), while those directed at Kemal Kılıçdaroğlu (opposition leader) remained significantly lower (9.2%), concretizing the power imbalance between the ruling party and opposition in Türkiye’s political landscape. Furthermore, the surge in sarcastic expressions during critical periods, such as economic crises and coup attempts, confirms that this rhetoric functions as a collective response mechanism. The results suggest that sarcasm in digital platforms could serve as an early warning signal for public sentiment under conditions of democratic decline. Future studies across diverse countries and larger datasets could deepen our understanding of this linguistic strategy’s role in political discourse, offering new insights into sociopolitical psychology.