Reidentifying Collective Entities after Splitting and Merger in Generative Relational Models

Collective splitting and merger distribute members, practices, commitments, resources, and recognized standing among different successors. This paper develops a method for reidentifying organized collectives through such changes within a generative relational framework. It distinguishes source records, normalized appearances, typed transmission, identity relations, and the assignment of subsequent consequences. Positive and negative identity premises determine families of globally admissible equivalences; their secure and possible information remains distinct from a complete historical finding. Conditional derivations examine consistency, conflicting premise repairs, uncertain normalization, temporal aggregation, hidden appearances, and the joint information lost by pairwise summaries. Membership, operation, undertaking, and institutional criteria receive separate justification and criticism. Evidence symmetry constrains unique selection, while adjacent matching and path examples expose the effects of objective, horizon, and score semantics. Rule-changing histories and grounded authority distinguish realized acts from their institutional warrant. Fallible documentary queries separate recovery of a target from recovery of an entire history and make shared-source failures explicit. Constructed merger, dormancy, allocation, and revision dossiers connect the results to complete inquiries. A bounded comparison of official Manchester documents distinguishes formation, heritage, membership carryover, and ceremony. Further analysis addresses historical action, undertaking succession, criterion validation, and exact integration of local findings. The contribution is an integrated evidential procedure with explicit antecedents and inspectable assumptions. Its mathematical results are conditional; broader empirical adequacy and a uniquely privileged numerical-identity criterion remain open.