A Candidate Decipherment of the Voynich Manuscript: Evidence for a Spoken Elu-Sinhala Pharmaceutical Register (V4)

Version 4 of the candidate decipherment of the Voynich Manuscript, identifying it as a 15th-century Sri Lankan Elu-Sinhala pharmaceutical text. This version supersedes V3 (Zenodo DOI 10.5281/zenodo.20072618, 2026-05-07). Major new findings in V4: V21 meaning corrections (26 word types, 5,307 tokens) from DPD, Clough, and BM cross-reference; COSMO decoder extended to 97.6% assignment coverage with 13/27 Pali nakkhatta identifications (8 BM-sequence supported, 3 confirmed); three internal structure tests independent of English glosses (section classifier p=0.0099; KALPANA enrichment OR=8.11; q-/ch- allomorphy OR=32.81); locked-anchor language comparison (Sarartha 66.67% vs Ortus Sanitatis 15.15%; all Siddha/Unani/Tamil/European alternatives 0%); three-stage blind botanical review with direct convergence on Papaver somniferum (f33r) and partial convergence on Nelumbo nucifera (f28v); plant section operating system confirmed as pharmaceutical ingredient register (87.9% formula-like; species-level claims downgraded to candidates); Team B validation suite; decoder frozen V17+V21/V27 with tiered lexicon. P(wrong) ~7-10%. Licensing: Original content CC-BY-4.0. Bundled third-party corpora retain their own licenses; see LICENSE.