War in Digital Time: A Computational Analysis of Cyclical War Discourse on Social Media

Traditional scholarship in war discourse assumes a linear progression of public engagement that mirrors the physical evolution of armed conflict: from outbreak through sustained combat to eventual resolution. This article challenges this paradigm by demonstrating that contemporary war discourse, particularly in digital spaces, operates through dynamic, recurring cycles rather than along a linear trajectory. Drawing from an unprecedented dataset of over thirty-four million tweets and retweets about the Syrian conflict collected since 2011, I reveal how public discourse at the popular level repeatedly cycles through interpretative phases, constructing and reconstructing frames for understanding major political and military events. The research employs a sophisticated methodological approach, utilizing a supervised machine-learning framework with a deep neural network (DNN) classifier and embedding to analyze the emotional content of war-related social media discourse. I demonstrate that these discursive cycles manifest through emotionally-charged and often violent rhetoric during specific periods, continuously reshaping public understanding of key developments in the conflict. This method enables the classification of daily emotional patterns through the aggregation of emotionally-laden tweets, which are then mapped against significant events in the conflict. This novel analytical framework permits large-scale examination of the relationship between online discourse and concrete events, revealing how digital war discourse evolves through repetitive cycles in response to developments on the ground. The implications of this study extends beyond the Syrian conflict and makes significant contributions to both theoretical and methodological approaches to analyzing war discourse. By combining advanced computational methods with traditional discourse analysis, it establishes a robust framework for understanding the evolution of public engagement with conflict in digital spaces. The evidence suggests that emotional responses and interpretative frameworks undergo continuous renewal and reshaping, creating a more dynamic and complex pattern of public engagement than previously recognized.