From subtext to prejudice: computational detection of implicit antisemitic tropes in online narratives

This paper investigates the transformation of media discourse surrounding Israel in the aftermath of the events of October 7, 2023, and the subsequent military escalation between Israel and Hamas. Using a corpus of textual data extracted from social media (the former Twitter platform, now X), it has been traced how narratives have shifted from a broader framing of the “Israeli–Palestinian conflict” to a targeted rhetoric that increasingly isolates Israel as a singular antagonist. The analysis reveals that public discourse has not only intensified in negativity toward Israel, but also frequently revives long-standing antisemitic tropes under the guise of political critique. Two main analytical dimensions guide the study. First, it is demonstrated that discourse on Israel is often conflated with discourse on the Jewish community as a whole, despite the socio-cultural and religious heterogeneity of Israeli society. This conflation obscures the internal plurality of Israel, reinforcing reductive stereotypes. Second, it is argued that social media rhetoric tends to blur the distinction between Israel as a state actor and Hamas as a non-state terrorist organization. In doing so, it frames the conflict as a generalized “Israel vs. Palestine” dichotomy, thereby oversimplifying the nature of the hostilities and fueling polarizing, hostile speech. By grounding the analysis in empirical data, the study highlights how social media acts as an amplifier of both contemporary geopolitical narratives and inherited prejudices. The findings underscore the need for a more nuanced understanding of the discourse on Israel, one that avoids collapsing complex realities into binary oppositions and resists the resurgence of antisemitic sentiment in digital arenas.