The rite keeps copying, the classroom keeps a relic

Claim (verbatim)

Indian Buddhist literature survives in Sanskrit through two exile channels with opposite physics, and the Tibetan Tanjur is the register against which both can be read. After the monastic universities fell around 1200, the tantric corpus stayed liturgically alive in the Kathmandu Valley's Vajrayāna, and living ritual demands fresh legible copies — so the great tantras and sādhana collections (Hevajra, Guhyasamāja, the Cakrasaṃvara corpus, the Sādhanamālā) kept moving down a Nepalese recopying treadmill into the nineteenth century. The epistemological curriculum — Dharmakīrti's seven treatises and their commentaries — lost its Indian classroom entirely and survives in Sanskrit only as relics: old palm-leaf carried to Tibet, kept as sacred objects at Sakya, Ngor, and Zhalu, photographed by Rāhula Sāṅkṛtyāyana in the 1930s and now surfacing in diplomatic editions. Same language, same religion, opposite survival shapes — lineage versus tomb. Prediction: for works of each class with any extant Sanskrit, the witness profiles will differ categorically — the named tantric works will show multi-witness Nepalese series including copies made after 1600, while Dharmakīrti's seven treatises will show single or few witnesses, none copied after 1500, with the earliest witnesses preserved in Tibetan monastic collections rather than Nepal (primary clause: the median witness count in the Nepalese microfilm record for the named tantric works is at least fivefold that of Dharmakīrti's seven treatises; the verdict follows it). Kill: the NGMCP/NGMPP online catalogue of the Nepal-German Manuscript Preservation Project (Hamburg), Rāhula Sāṅkṛtyāyana's Tibet manuscript lists in the Journal of the Bihar and Orissa Research Society 21-24 (1935-1938), the Sanskrit Texts from the Tibetan Autonomous Region editions (China Tibetology Research Center and Austrian Academy of Sciences), and the Tōhoku Complete Catalogue of the Tibetan Buddhist Canons (Sendai, 1934) for the Tanjur sections.

Prediction clause (verbatim)

Prediction: for works of each class with any extant Sanskrit, the witness profiles will differ categorically — the named tantric works will show multi-witness Nepalese series including copies made after 1600, while Dharmakīrti's seven treatises will show single or few witnesses, none copied after 1500, with the earliest witnesses preserved in Tibetan monastic collections rather than Nepal (primary clause: the median witness count in the Nepalese microfilm record for the named tantric works is at least fivefold that of Dharmakīrti's seven treatises; the verdict follows it).

Kill-dataset (verbatim)

Kill: the NGMCP/NGMPP online catalogue of the Nepal-German Manuscript Preservation Project (Hamburg), Rāhula Sāṅkṛtyāyana's Tibet manuscript lists in the Journal of the Bihar and Orissa Research Society 21-24 (1935-1938), the Sanskrit Texts from the Tibetan Autonomous Region editions (China Tibetology Research Center and Austrian Academy of Sciences), and the Tōhoku Complete Catalogue of the Tibetan Buddhist Canons (Sendai, 1934) for the Tanjur sections.

Novelty / leakage triage

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

The two-channel survival story is standard: Steinkellner's 'A Tale of Leaves' describes the pramana palm-leaves kept as relic-objects in Tibetan monasteries, and Sankrityayana's lists plus the NGMPP literature document the Nepalese liturgical recopying of the tantric corpus into the nineteenth century. No one has computed the witness-count differential between the two classes (the fivefold median ratio over the Nepalese microfilm record).

• E. Steinkellner, A Tale of Leaves: On Sanskrit Manuscripts in Tibet, their Past and their Future (2003 Gonda Lecture, Royal Netherlands Academy, Amsterdam, 2004)
• R. Sankrityayana, 'Sanskrit Palm-leaf Mss. in Tibet', Journal of the Bihar and Orissa Research Society 21 (1935) and 23 (1937)

Predictions

No prediction registered yet.