Exploring Turkish narratives on Syrians: An analysis of Twitter discourse (2010-2022)

This dissertation investigates evolving patterns of public discourse surrounding the Syrian conflict and refugee crisis in Turkey, analyzing more than a decade of Turkish language Twitter conversations (2010–2022). It reveals a remarkable 1,200-fold increase in tweet volume, reflecting a shift from episodic attention to a broader, emotionally driven engagement with Syria-related topics. To capture the dynamics of this discourse, the study develops a custom-built Emotion Recognition Model (ERM) by fine-tuning BERTurk on Turkish emotional datasets, marking one of the first emotion recognition tools tailored for Turkish political discourse. Through a multi-method approach that integrates ERMs, topic modeling (BERTopic and LDA), and semantic network analysis, the research uncovers how digital publics frame conflict, identity, and migration over time. A comparative analysis of original and de-duplicated tweet datasets highlights how coordinated amplification distorts the perceived spontaneity of public sentiment. Focusing on four critical years (2014, 2016, 2017, and 2022), the study identifies distinct communicative modes: emotional amplification, political signaling, silence under pressure, and the emergence of xenophobic narratives. The findings demonstrate that Twitter operates not merely as a channel for opinion, but as a platform where affective language constructs shared meaning. Conceptually, the dissertation draws on Social Constructionism, Framing Theory, and Affective Publics to argue that emotionally charged narratives shape digital engagement more than rational debate. By contrasting Twitter data with Google Trends, it introduces the notion that Turkish users "feel through Twitter, think through Google," offering an original lens on how affect and attention diverge across platforms. This work advances the study of digital political communication by combining novel tools, large-scale data, and theory-driven analysis to illuminate how online discourse constructs public understanding of conflict and belonging.