The Computational Models of Narrative (CMN) workshop series is dedicated to advancing the computationally-grounded scientific study of narrative, a crucial aspect of human experience used for communication, persuasion, explanation, and entertainment. Narrative, or storytelling, is a symbolic activity that imitates human actions through emplotment, representing discordant events through concordance. From myths to histories, narratives are ubiquitous across time, making understanding narrative intelligence essential to comprehending human intelligence. Narrative studies, aka narratology, has its root in Aristotle's Poetics, thriving in the wake of the French New Rhetoric, where Todorov first coined "narratology" in 1969 to theorize narrative grammar based on structuralist linguistics. Despite criticism for its formalism and reduction, this characteristic enables the intersection of computer science and poetics, providing various structures for computational modeling.
Computer scientists have long tapped into the three-act structure, Freitag's pyramid, Propp's morphology, and Campbell's or Vogler's hero's journey. Large Language Models (LLMs) boast their breakthrough in generating narratives, betraying traces of the structures mentioned above. Systems for narrative analysis and production are increasingly embedded in devices and processes, influencing decision-making in venues as diverse as politics, economics, intelligence, and cultural production. In order to appreciate this influence, it is becoming increasingly clear that research must address the technical implementation of narrative systems, the theoretical bases of these frameworks, and our general understanding of narrative at multiple levels, from the philosophical and cognitive impact of narratives to our ability to model narrative responses computationally.