EVIDENT: Measuring Semantic Convergence in Governed Ecosystems

Semantic interoperability has traditionally been treated as a property of semantic artifacts, measured indirectly through publication, adoption, or institutional endorsement. Despite decades of standardisation efforts, ontology engineering methodologies, and semantic governance programmes, competing conceptualisations continue to proliferate across domains, leaving knowledge modellers without a principled design-time mechanism for selecting the semantic artifacts most likely to maximise interoperability within their operational context. This position paper argues that semantic interoperability should be understood as an emergent property of convergence among identifiable parties operating within a bounded representation scope, rather than as a property of semantic artifacts in isolation. To explore the implications of this position, we introduce EVIDENT and defines interoperability relative to an identifiable party set (P), a bounded representation scope (R), and a class of formally grounded semantic artifacts (A). Within this context, the framework derives three machine-computable evidence signals: convergence of definition through formal architectural commitments (Cdf), convergence of definition through semantic alignment (Cda), and convergence of adoption through operational deployment (Ca). These signals are combined through an Evidence Ordering Function that produces a queryable partial ordering over competing artifacts, while a governance footprint separates institutional authority from empirical authoritativeness. The framework is intended to provide knowledge engineers and autonomous agents with design-time evidence for evaluating semantic modelling choices before deployment. Through a cross-border pension entitlement scenario, the paper illustrates how EVIDENT identifies convergence patterns among peer organisations and supports artifact selection based on observable ecosystem behaviour rather than mandate, visibility, or philosophical precommitment. The framework proposes a mechanism that can help make the cost of semantic divergence visible, attributable, and computable within governed ecosystems.