Recitation beats revision

Claim (verbatim)

Recitation beats revision: in Chinese Buddhism, translation quality lost to liturgical installation. Once a version of a scripture was embedded in recitation practice, philologically superior retranslations — including the state-sponsored Xuanzang corpus — could not displace it in quotation; new versions won uptake only inside the doctrinal lineage that produced them. The effective canon was fixed at the level of memorized sound, so version choice in citation should be predicted by the citing author's school lineage, not by the availability or age of better translations — a topological claim about uptake, not the familiar observation that Kumarajiva stayed popular.

Prediction clause (verbatim)

In CBETA, collect identifiable quotations of the Diamond Sutra in Chinese works composed after 800, when six competing translations existed. Primary clause: outside Faxiang/Yogacara-lineage works, at least 90 percent of quotations follow Kumarajiva's version rather than any later translation; the verdict follows this clause. Secondary clause: within Faxiang-lineage commentaries the Xuanzang-corpus versions reach at least parity with Kumarajiva, isolating the mechanism as school-liturgical rather than chronological.

Kill-dataset (verbatim)

Kill: the CBETA machine-readable Chinese Buddhist canon (full Taisho text, all competing Diamond Sutra translations included), searchable for quotation wording.

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: Leaked (already exists in the literature).

The item's primary clause -- that outside the Faxiang/Yogacara lineage >=90% of post-800 Diamond Sutra quotations follow Kumarajiva's version despite five later, more literal translations including Xuanzang's -- restates a textbook fact: Kumarajiva's 402 CE rendering remained the dominant, most-chanted version and was not displaced by the philologically superior later translations. The verdict follows this clause, so it tracks an established observation the item itself concedes is familiar. The genuinely fresh content -- version choice predicted by the citer's school lineage, with Faxiang commentaries reaching parity with the Xuanzang corpus -- is relegated to the secondary clause and does not carry the verdict.

• Tingting Mi, 'A Study of the Diamond Sutra and its Different Versions' (Atlantis Press) — Kumarajiva's version dominant across six translations, including Xuanzang's, in later uptake
• CBETA machine-readable Chinese Buddhist canon (all competing Diamond Sutra translations) — The kill dataset for lineage-conditioned quotation-wording analysis

Predictions

No prediction registered yet.