Structural Identification of the Voynich Manuscript Token Sequence

This project presents a completed, preregistered, and reproducible structural analysis of the Voynich Manuscript symbol sequence. The work demonstrates that the token stream is consistent with generation by a single continuous one-dimensional stochastic latent process with noisy readout, continuous across all manuscript sections. The analysis is structural only. It does not claim translation, decipherment, language identification, or semantic interpretation. Five non-identifiable properties are proven to be information-theoretically inaccessible from symbols alone. Phases 81-89 constitute a closed body of results. Phases 90A-90C provide access and measurement artifacts only. Phase 91 is preregistered and must not be executed until qualifying external physical data exist. All claims, limitations, and governance constraints are documented in the accompanying release declaration.