The Voynich Manuscript as a Structural Loop System

This paper does not present a decipherment, translation, hidden language, cipher key, or historical identification of the Voynich Manuscript. It proposes a structure-first analytical framework — the Voynich Multi-Artifact Framework (VMAF) — for examining the manuscript prior to semantic interpretation.

The study models the manuscript as a multi-domain structural artifact rather than as a purely linear text. It evaluates recurrence, positional stability, relational consistency, cross-context persistence, loop-based validation, manuscript-scale continuity, and visual–topological organization. Through this procedure, the paper identifies a stable CORE layer of recurring structural units and tests whether those units participate in constrained return patterns across textual, visual, and spatial domains.

The analysis also examines the Rosettes foldout (f85v–f86r) as a visual–topological structure, focusing on localized structural regimes, routing behavior, bridge zones, and network-like organization. Controlled projection is introduced not as a decoding method, but as a bounded comparative procedure for testing whether internally validated structural roles remain admissible under preserved constraints.

The central claim is limited: the manuscript remains unresolved in meaning, but its internal organization can be studied as a reproducible, constraint-governed structural system. The study therefore reframes the Voynich problem as one of structural validation prior to semantic explanation.

A separate methodological supplement is available here:
https://zenodo.org/records/20622484

The supplement clarifies the VMAF experimental protocol, including GPT-assisted Builder / Evaluator / Supervisor role separation, EVA-based textual testing, Yale/Beinecke visual-only analysis, held-out validation, frozen outputs, failure conditions, and requirements for independent critique or replication.