Relational Search Operator Specification 0.1: Frozen Method, Validation Record, and Prospective Test Protocol

Abstract

Relational Search Operator Specification 0.1 is the first frozen normative checkpoint of an explicit reasoning/search procedure for organizing evidence around scoped relations, witnessed boundary crossings, materially different candidate structures, consequence and residue, contradiction and valid absence, rejection conditions, and best-next-question handoff. The release preserves the operator together with its pre-freeze validation record, authorised claims and limitations, provenance, evidence index, frozen comparative results, and a prospective blind Human vs Strong Baseline vs Relational Search vs external-reality test protocol that had not yet been run when version 0.1 was frozen. Across the pre-freeze benchmarks, Relational Search showed robust and calibrated handling of functional, temporal, identity, provenance, contradiction, resolution, authority, causal, and correlation-only cases. Direct controlled comparisons did not demonstrate material superiority over a strongest general analytical baseline: CP1 and CP2 were primary-score ties; CP3 is retained as a confounded comparison because its baseline instructions were inadvertently relationally primed; and CP4 was a valid clean 20/20 versus 20/20 tie. The release therefore presents Relational Search as an explicit and recoverable reasoning/audit architecture, not as a demonstrated superior-intelligence or automatic-discovery system. This public freeze is intended to make subsequent validation prospective and auditable. Future successful applications, proofs, experiments, or domain discoveries retain their own discovery credit; they should not be attributed to this specification merely because the method helped formulate a useful question or relation.