The curse is a trade

Claim (verbatim)

The roughly 130 lead prayers from the sacred spring at Bath are often imagined as furious bathers scratching their own maledictions, but the tablets' palaeography permits a head count of hands, and hands are the tell: personal writing predicts nearly as many hands as tablets, while a scribal service at the temple gate predicts few hands serving many clients, plus formularies repeated verbatim with slots for names. The same test runs at the rural shrine of Uley. If hand concentration holds, cursing was a writing industry — the precinct sold documents to the aggrieved as notaries sold contracts — and curse tablets measure demand for writing services, not the distribution of writing skill.

Prediction clause (verbatim)

In the published Bath (Tab. Sulis) and Uley corpora as presented in Roman Inscriptions of Britain online, distinct identified hands number fewer than half the legible tablets, and at least three formulaic templates recur across tablets in different clients' names. Primary clause: the hands-to-tablets ratio below 0.5 — a hand-count statistic.

Kill-dataset (verbatim)

Roman Inscriptions of Britain online (Tab. Sulis and Uley curse-tablet corpora).

Novelty / leakage triage

already answered in the literature

Tomlin's Tab. Sulis edition ran precisely this test — detailed hand tabulations across the Bath corpus assessing professional-scribe vs personal authorship — and concluded 'amateurs at work': hands are nearly as numerous as tablets, with only rare same-hand pairs (e.g. Tab. Sulis 95/96) and formula variation telling against a scribal service. The hand-count-as-industry-detector connection is published, with the answer.

Predictions

No prediction registered yet.