Extending the Boundaries of Paracrises: A Computational Perspective on Narrating Stalemates Through eWOM in Electing Mr. Speaker on X

ABSTRACT Social media play crucial dialogic and strategic roles during paracrisis. Responding to a prior call to extend agenda‐setting theory to study paracrises, this manuscript builds on the literature by moving beyond the dominant focus on for‐profit organisations in public relations and crisis studies into the political sphere. Adopting unsupervised machine learning and qualitative methods, this study used a comprehensive dataset of 2451 user‐generated comments on X during the 2023 U.S. House of Representatives Speaker election to contribute to the broader discourses on paracrises. Findings from topic modelling revealed that the agenda narrative focused on House proceedings and conflicts , the GOP's internal struggle, a leadership crisis and Democratic alternatives , as well as shared political interests and ideologies . In addition, a combination of sentiment analysis and sentence mapping highlighted positive user comments focused on the theatricalisation of serious politics and on appreciation for Democratic comportment . Conversely, negative comments emphasised attacks on figures/party, criticism of legislative dysfunction , and mismanagement of resources . This study first advances a definition for political paracrisis and highlights the profound contribution of these findings to the praxis of crisis studies.