A kingdom governed in a borrowed prakrit

Claim (verbatim)

In the sands south of the Taklamakan, at Niya and Loulan and Endere, Aurel Stein dug up close to a thousand documents on wood and leather written in Kharoshthi script and Gandhari Prakrit — the language and script of the far northwest of the subcontinent, here running the chancery of the third-century Shanshan (Kroraina) kingdom two thousand kilometres from where that Prakrit was a mother tongue. It is a near-complete administrative literature of a lost desert state: orders, contracts, tax and corvee records, letters under seal. And it is almost the only thing left of Shanshan's writing; the language died there, the kingdom was abandoned, and the local Iranian and other vernaculars behind the Gandhari facade barely surface at all. The mechanism is a prestige-transplant filter: a conquest or missionary language installed as the medium of record leaves its administrative deposit and buries the substrate speech beneath it, so what the sand preserved is the imported officialese, while the everyday languages of the oasis are legible only as loanwords and names inside the Prakrit. The corpus therefore over-represents the chancery and effaces the vernacular that surrounded it. Prediction: classifying the Niya-Loulan Kharoshthi corpus by function, administrative-legal-epistolary documents will exceed 90% of items and literary or religious texts fall under 5%, while direct traces of the oasis's non-Gandhari vernacular appear only as embedded loanwords and onomastics, not as free-standing texts (primary clause: the over-90% administrative share with no free-standing vernacular text; the verdict follows it). Kill: A. M. Boyer, E. J. Rapson and E. Senart, Kharoshthi Inscriptions Discovered by Sir Aurel Stein in Chinese Turkestan (1920-1929) and T. Burrow, A Translation of the Kharoshthi Documents from Chinese Turkestan (1940), now machine-readable as the Niya documents in the Corpus of Kharoshthi Documents at gandhari.org.

Prediction clause (verbatim)

Prediction: classifying the Niya-Loulan Kharoshthi corpus by function, administrative-legal-epistolary documents will exceed 90% of items and literary or religious texts fall under 5%, while direct traces of the oasis's non-Gandhari vernacular appear only as embedded loanwords and onomastics, not as free-standing texts (primary clause: the over-90% administrative share with no free-standing vernacular text; the verdict follows it).

Kill-dataset (verbatim)

Kill: A. M. Boyer, E. J. Rapson and E. Senart, Kharoshthi Inscriptions Discovered by Sir Aurel Stein in Chinese Turkestan (1920-1929) and T. Burrow, A Translation of the Kharoshthi Documents from Chinese Turkestan (1940), now machine-readable as the Niya documents in the Corpus of Kharoshthi Documents at gandhari.org.

Novelty / leakage triage

already answered in the literature

The Niya-Loulan Kharoshthi corpus is universally characterised as the administrative, legal, and epistolary paperwork of the Kroraina/Shanshan chancery - orders, contracts, tax and corvee records, sealed letters - with only a negligible literary or religious remainder, in Boyer-Rapson-Senart's edition, Burrow's translation, and the machine-readable CKD. The over-90% administrative share, and the survival of the oasis's non-Gandhari vernacular only as embedded loanwords and names, are the stated structure of the corpus, not an un-run audit. The primary clause's answer is already in the published record.

• A. M. Boyer, E. J. Rapson & E. Senart, Kharoshthi Inscriptions Discovered by Sir Aurel Stein in Chinese Turkestan, 3 parts (Oxford: Clarendon, 1920-1929)
• T. Burrow, A Translation of the Kharoshthi Documents from Chinese Turkestan (London: Royal Asiatic Society, 1940)
• Corpus of Kharoshthi Documents (CKD), gandhari.org (S. Baums & A. Glass)

Predictions

No prediction registered yet.