A Functional Decipherment of the Indus Valley Script Without Phonetic Assumptions — 90.5% Token Coverage Across 2,543 Inscriptions

Public priority claim for a functional decipherment of the Indus Valley script using the Molina Methodology. Four structural layers identified without phonetic assumptions: (1) administrative grammar — opener sign glyph 740 (p = 5.46 × 10⁻²⁸⁵) and five closer signs (all p < 10⁻¹⁵); (2) mathematical layer — quantity signs preceding classifier sign at 86–100% (all p < 0.00001); (3) iconographic anchoring — 116 sign types assigned via animal co-occurrence analysis; (4) proper noun residual — 9.5% of tokens, 418 unique types, statistical signature of names. Total functional coverage: 90.5% of 11,280 sign tokens across 2,543 inscriptions from 31 sites. A bilingual text would assist phonetic assignment of the 9.5% proper noun layer only. Full methodology withheld pending peer review. This deposit extends Molina (2026a) — structural falsification of Sproat's non-linguistic hypothesis, deposited March 16, 2026. SHA-256 of full private paper: 65393687bd54cbd3cf20ef4eeda1c057f639fe686a6467636ff5798df95817ce. SHA-256 of private methodology: 992bec189397bd59c2033c0c486f5491b4ea0c3faf40c988116aef4d926bf861. Data: CISI (Parpola et al.), yajnadevam/indus-website (GPL-3.0). All results independently reproducible.