Recursive Semantic Anchoring in ISO 639:2023: A Structural Extension to ISO/TC 37 Frameworks

Abstract

ISO 639:2023 unifies the ISO language-code family and introduces contextual metadata, but it lacks a machine-native mechanism for handling dialectal drift and creole mixtures. We propose a formalisation of recursive semantic anchoring, attaching to every language entity a family of fixed-point operators _n,m that model bounded semantic drift via the relation _n,m() = (), where () is a drift vector in a latent semantic manifold. The base anchor _0,0 recovers the canonical ISO 639:2023 identity, whereas _99,9 marks the maximal drift state that triggers a deterministic fallback. Using category theory, we treat the operators _n,m as morphisms and drift vectors as arrows in a category DriftLang. A functor : DriftLang AnchorLang maps every drifted object to its unique anchor and proves convergence. We provide an RDF/Turtle schema (BaseLanguage, DriftedLanguage, ResolvedAnchor) and worked examples -- e.g., _8,4 (Standard Mandarin) versus _8,7 (a colloquial variant), and _1,7 for Nigerian Pidgin anchored to English. Experiments with transformer models show higher accuracy in language identification and translation on noisy or code-switched input when the -indices are used to guide fallback routing. The framework is compatible with ISO/TC 37 and provides an AI-tractable, drift-aware semantic layer for future standards.

Tasks

Reproductions