The jar remembers what the canon forgot

Claim (verbatim)

The Gāndhārī birch-bark scrolls — the British Library collection acquired in 1994, then the Senior, Bajaur, Split, and kindred collections — are the oldest manuscripts of South Asia, written around the turn of the common era and the two centuries following, and they reached us by a route unlike any canon: deposited in jars and stūpa precincts, some apparently retired when worn, the Senior group seemingly commissioned for interment as a set. A deposit is not a curriculum. It samples what a working monastic community actually had on hand, with none of the canon-formation filter that decided, centuries later and elsewhere, what deserved recopying — so the composition of the sample directly estimates how much of the ambient textual population the surviving canons dropped. The corpus is now catalogued text by text, with parallels in the Pali, Sanskrit, Chinese, and Tibetan canons recorded where they exist, and Salomon's surveys already signal that the commentaries and the local avadāna collections largely have none. Prediction: classifying the literary manuscript texts in the Catalog of Gāndhārī Texts (excluding documents and inscriptions), at least one-third will have no identified parallel in any surviving canon, and the unparalleled share will run highest in the commentarial, scholastic, and avadāna genres — the genres canon-formation most freely discarded (primary clause: the one-third unparalleled floor; the verdict follows it). Exact computation: tabulate the catalogue's parallel fields by genre class and compute shares. Kill: the Catalog of Gāndhārī Texts at gandhari.org (Baums and Glass), with Richard Salomon's Ancient Buddhist Scrolls from Gandhāra (Seattle, 1999) and The Buddhist Literature of Ancient Gandhāra (2018) as the interpretive controls.

Prediction clause (verbatim)

Prediction: classifying the literary manuscript texts in the Catalog of Gāndhārī Texts (excluding documents and inscriptions), at least one-third will have no identified parallel in any surviving canon, and the unparalleled share will run highest in the commentarial, scholastic, and avadāna genres — the genres canon-formation most freely discarded (primary clause: the one-third unparalleled floor; the verdict follows it). Exact computation: tabulate the catalogue's parallel fields by genre class and compute shares.

Kill-dataset (verbatim)

Kill: the Catalog of Gāndhārī Texts at gandhari.org (Baums and Glass), with Richard Salomon's Ancient Buddhist Scrolls from Gandhāra (Seattle, 1999) and The Buddhist Literature of Ancient Gandhāra (2018) as the interpretive controls.

Novelty / leakage triage

already answered in the literature

The corpus's chief editor has already published the result: Salomon states that the majority of the Gandhari texts - the avadanas and the commentarial/scholastic texts above all - have no parallels in the previously known Buddhist literatures, which asserts both the one-third floor and the genre gradient of the primary clause; only the formal tabulation of gandhari.org's parallel fields remains to be done.

• R. Salomon, 'Gandhari, Buddhist Literature in', in Encyclopedia of Buddhism, ed. R.E. Buswell (New York, 2004)
• R. Salomon, The Buddhist Literature of Ancient Gandhara (Somerville, 2018)

Predictions

No prediction registered yet.