#136: Structural Signals and Role Archetypes: The Emergence of Roles in a Distributed Epistemic Game

Roles in collaborative learning are often scripted or assigned; in distributed epistemic games (DEGs), however, they emerge through interaction over time. This study analyzes the first Polymath Project, a large-scale mathematical collaboration, to show how roles crystallize in practice. Using canonicalized epistemic move sequences, temporal distance metrics, and cross-method clustering, I identify two converging empirical signals. First, participant distance matrices display visible “plaid” structures—evidence that contributors gravitated into role-like repertoires even before clustering. Second, across five analytic pipelines, three participant groupings (‘consistent companionships’) persisted: Proof Builders, Paired Refiners, and Process Integrators. Their recurrence across methods supports viewing them as durable role archetypes. Conceptually, the findings show that roles in DEGs are emergent and self-organizing; methodologically, they demonstrate how structural and reflexive analysis (via AIRR) can jointly surface roles-in-practice, advancing methods for studying collective intelligence and informing the design of systems that support dynamic role ecologies.