Indus Signs as Merchant Marks: Corpus Structure, Context, and Viability

The inscriptions of the Indus Valley Civilisation are extremely short (modal 4–5 signs) and occur on portable administrative media that co-occur with weights and seals. We evaluate the corpus as a merchant-mark system supporting trade administration rather than extended phonetic writing. Using aggregated catalogues and provenienced contexts, we assemble corpus statistics, layout regularities, and duplication rates, and we test a large-format case from Dholavira. In that case we propose a working identification of a complete animal-drawn cart assembled sign-by-sign; this exemplar, together with emblem-plus-label layout, low internal repetition, exact string duplication on sealings, and spatial concentration at gates, workshops, and storerooms, functions as positive evidence for compact commercial identifiers and, by basic elimination, as negative evidence for continuous running text. We derive predictions linking recurrent signs to metrology and spatial context and provide a cautious lexicon of iconographic readings for high-frequency signs. The model does not exclude phonetic values; it argues that the dominant preserved use is administrative labelling within a high-throughput Bronze Age economy