From lexical equality to receiving entries in the Voynich Manuscript

A repeated transcription string can identify a candidate inscription without establishing a historical address, an entry's owner or an instruction to consult it. This study reconciles all 989 links in a previously computed herbal-to-pharmaceutical equality graph, retaining every one of its 1,257 receiving references. In the fixed transcription, 828 references reach complete, certain receiving labels and 429 reach components or other noncomplete forms. Only two references have a complete source label: receiver completeness cannot be transferred to the source side. Duplicate references and shared images are retained as dependencies. A prospective stratified selection leads to two previously unadjudicated ownership questions. Native images locate the inscriptions and support local botanical associations, while competing owners and larger entry boundaries remain explicit. Comparison with retained transcription versions reveals receiving-spelling disagreement in one case and source-segmentation dependence in both. The result is a fully typed reference ledger and two bounded source tests. It identifies what the matches can locate and exactly what is still required to turn them into a stable lexical address or historical operation. [B01–B10] This preprint is one of four companion works in the Voynich Reader Studies series. The shared release archive contains the companion manuscripts, reader guides, source matrices, figures and reproducible evidence. Scientific qualifications remain attached to their respective tests.