The unicorn monopoly

Claim (verbatim)

The unicorn monopoly. Joins the concentration indices of industrial organization — the tools economists use to tell competitive markets from managed ones — to Indus glyptic. Stamp seals were the emblems of whoever ran Indus commerce, and if their animal motifs were the competing badges of independent houses or cities, the motif mix should vary from town to town as local players rose and fell. Instead, the claim runs, the unicorn behaves like a single managed franchise with an enforced share: its fraction of stamp seals should sit in the same narrow band — around two-thirds — at every major site, with a cross-site spread of only a few percentage points and near-identical motif concentration indices, held constant across a million square kilometres. Contemporaneous Mesopotamian glyptic, where cities genuinely differed, supplies the contrast: there motif shares swing widely from city to city, so the Indus uniformity is no generic property of seal-using societies.

Prediction clause (verbatim)

In the Corpus of Indus Seals and Inscriptions, the unicorn motif's share of stamp seals lies between 60% and 75% at every major site (Mohenjo-daro, Harappa, Chanhu-daro, Lothal, Kalibangan, Dholavira) with a cross-site spread under 8 percentage points, and site-level motif Herfindahl indices agree within +/-15%; by contrast, motif shares in contemporaneous Mesopotamian glyptic vary across cities by more than 25 points, so the Indus uniformity is not a generic property of seal-using societies.

Kill-dataset (verbatim)

Kill: CISI motif-by-site tabulations, with Mesopotamian glyptic catalogues as the contrast set. Unicorn shares spreading over more than 20 percentage points across major Indus sites kills it.

Novelty / leakage triage

anticipated in the literature — this exact test has never been run

Per-site unicorn shares circulate in the literature-adjacent record (reported ~60% Mohenjo-daro vs ~46% Harappa — a ~14-point spread that would FALL OUTSIDE the conjecture's predicted <8-point band, and below its 60-75% floor at Harappa), and quantitative unicorn-seal studies are active (2023 multivariate stylistic study; seal network analyses). The Herfindahl-by-site computation with a Mesopotamian contrast was not located. Any registration must confront the informal figures' contrary indication first — this conjecture may be born falsified.

• Harappa.com Indus-seal overview (per-site unicorn shares) — Existing figures sit outside the predicted band
• 'Unicorns and seals: New multivariate approaches...', JAS Reports (2023) — Active quantitative treatment of the corpus

Predictions

No prediction registered yet.