This review draft presents a preliminary administrative-functional reading of eight Indus/Harappan seals and seal fragments. It does not claim a final decipherment of the Indus script and does not propose confirmed ancient phonetic values. The analysis separates three layers: (i) a high-confidence functional layer treating the objects as administrative or identity seals; (ii) a medium-confidence structural layer interpreting animal emblems, stands, sign rows, class markers, vessel signs and terminal signs as components of a short administrative formula; and (iii) a low-confidence controlled phonetic-label layer used only as computational notation for testing sign sequences.The sample is intentionally small and serves as a pilot for method development and expert feedback. The strongest anchor is the terminal cup/jar class (Mahadevan/Yadav M328 and M342 families). Phonetic labels (HA, U, KI, GAR, MI, JA, EN, KU, TA, KA, GA, NA, AG) are introduced solely as controlled computational notation and are not proposed as recovered ancient pronunciations. A follow-up version will include full sign-number mapping to Mahadevan, Wells and ICIT concordances.
Balevsky et al. (Sun,) studied this question.