Principlev1
Test bridge node validity by generating novel predictions or
Test bridge node validity by generating novel predictions or insights in both connected domains—if the connection produces only similarity recognition without actionable transfer, demote it to metaphor status.
Why This Is a Principle
Derives from Structural Correspondence in Mental Models (models have structural correspondence—real bridges preserve causal structure across domains), The performance of an agent is bounded by the accuracy of (agent performance bounded by world model accuracy—false bridges degrade models), and Learning occurs when outcomes differ from predictions, (learning requires prediction error—bridges must generate testable predictions). This operationalizes quality control for cross-domain connections.