Relational Joining of Separate Records: How Linking Accurate Fragments Creates a Claim None of Them Makes

This conceptual paper introduces Relational Joining of Separate Records, the operation by which links of identity, sequence, or reply between separately made records produce a composite claim. Holding the fragments fixed while changing their links changes what the combined account asserts. Accurate fragments therefore do not establish the accuracy of the relationship asserted between them. The paper distinguishes a relational join from mere co-presentation and assesses each link according to its own evidence. It asks how identity was established, how events were ordered, or what shows that one message responds to another. It also considers whether uncertainty attached to a link remains visible when the joined account circulates. The account relates the operation to Fellegi and Sunter’s theory of record linkage and examines the U.S. Commission on Civil Rights’ account of Florida’s voter-list matching before the 2000 election. The discussion distinguishes uncertain candidate matches, local verification practices, documented false identity links, and separate errors of classification. Part of the Generative Operations series within On Human Modeling, the paper offers questions for examining who creates, verifies, challenges, and repairs links. Its framework can inform human–AI research on whether relationships asserted in a synthesis are supported independently of the accuracy of its source fragments.