Iconoclasm is an event

Claim (verbatim)

Iconoclasm is an event. Rubbed-out faces in Persian and Arabic figural manuscripts are usually narrated as diffuse pious attrition accumulating across many readers. If that were so, defacement per manuscript would spread smoothly with use. But if defacement is typically one owner's single campaign through the book, the distribution is bimodal — books nearly untouched or nearly complete in their defacement — and defacement will ignore the wear gradients that genuine handling leaves.

Prediction clause (verbatim)

Primary clause (the verdict follows it): across at least 60 digitized figural manuscripts showing any face-defacement, the per-manuscript defaced-face fraction is bimodal, with at least 70 percent of manuscripts below 0.15 or above 0.85. Secondary clause: within manuscripts, defacement probability is uncorrelated with folio-level soiling and thumbing gradients, whereas genuine use-wear is front-loaded.

Kill-dataset (verbatim)

Kill: British Library and BnF digitized Persian and Arabic manuscripts, plus the Cambridge Shahnama Project, whose painting-level records note rubbed and repainted faces.

Novelty / leakage triage

provisional — model-triaged, shepherd review pending

A provisional first pass authored by the model (Opus), not yet confirmed by the shepherd. It carries the same dated-search requirements as an authoritative verdict but is excluded from every headline figure and cannot underwrite a prediction until a shepherd confirms it. Provisional reading: Adjacent (closely related prior work exists).

Finbarr Barry Flood's work on the defacement of figural manuscripts distinguishes exactly the item's two modes -- a single owner's 'bureaucratic, calculated, and protracted' campaign through a whole book (Evliya Celebi's 1655 Bitlis account of a buyer knifing the eyes/throats of every figure in a Shah-nama) versus diffuse ritual/pious rubbing -- so the episodic-campaign reading is documented. Not located: the item's distributional operationalization -- that the per-manuscript defaced-face fraction is BIMODAL (>= 70 percent of manuscripts below 0.15 or above 0.85) and that within manuscripts defacement is uncorrelated with folio-level soiling/thumbing gradients (unlike front-loaded genuine wear), across >= 60 digitized figural manuscripts. Both modes documented; the bimodality plus wear-decorrelation measurement is unrun.

• Finbarr Barry Flood, 'Between Cult and Culture: Bamiyan, Islamic Iconoclasm, and the Museum', The Art Bulletin 84.4 (2002) — Calculated, protracted single-owner defacement campaigns vs ritual/pious rubbing; the Evliya Celebi Shah-nama account.
• British Library and BnF digitized Persian/Arabic manuscripts and the Cambridge Shahnama Project (rubbed/repainted-face records) — The kill dataset for the per-manuscript defacement-fraction bimodality and wear-gradient decorrelation.

Predictions

No prediction registered yet.