This study proposes a structural interpretation of the Indus script as a procedural administrative encoding system embedded within seal iconography rather than a phonetic writing system designed to represent spoken language. Drawing on statistical observations of inscription length, positional constraints on signs, sign-frequency distributions, and iconographic patterns, the model reconstructs a functional grammar composed of institutional identifiers, merchant markers, commodity-domain signs, transaction modifiers, and closure operators. The analysis integrates iconographic elements—particularly the dominance of the unicorn motif and its associated device—into the administrative structure of the system. The model suggests that Indus seals functioned as compressed transaction certificates within a standardized trade network, explaining several otherwise puzzling features of the corpus, including short inscription length, the presence of large semantic sign families, and the absence of narrative texts. Rather than proposing phonetic decipherment, the framework focuses on the information architecture and administrative function of the script. The model remains compatible with multiple linguistic hypotheses including Dravidian and language-isolate scenarios. Several testable predictions are proposed concerning sign-position distributions, issuer clustering, and commodity-domain sign families within the Indus corpus.
Matthew Dominik (Thu,) studied this question.