The geographers saw what the archives never held

Claim (verbatim)

On the western steppe two literate-adjacent states — Khazaria and Volga Bulgaria — are richly described and barely self-attested. Almost everything narrated about them is outsider prose: Ibn Fadlan's eyewitness report of the 921-922 Baghdad embassy, the Balkhi-school geographers (al-Istakhri, Ibn Hawqal) on the Khazar capital and the Bulgar market, plus Hebrew and Byzantine notices. Native written output is close to nil. For the Volga Bulgars it is essentially a corpus of Islamic gravestones of the thirteenth-fourteenth centuries, in a distinctive Bulgar Turkic with Chuvash-type features — a language otherwise unrecorded, surfacing only to bury the dead. For the Khazars it is a contested handful of Hebrew-script letters (the Hasdai-Joseph correspondence, the Cambridge/Schechter document, the Kievan Letter) whose very authenticity has been argued for a century. The mechanism is that steppe polities administering through Sogdian, Arabic, or Hebrew intermediaries, and converting to religions whose prestige language was not their own, generated their record in other people's scripts and other people's archives; the native tongue reached writing only where a permanent medium and a personal motive coincided — the tombstone. Prediction: a census of primary written sources for Khazaria and Volga Bulgaria before 1300 will find that texts composed by outsiders outweigh securely native-authored texts by more than ten to one in word count, and that every securely native written witness in the Bulgar language is epigraphic and funerary rather than documentary or literary (primary clause: the all-epigraphic character of native Bulgar witness; a native Bulgar documentary or literary text falsifies it). Kill: the Volga Bulgar epitaph corpus in the editions of G. V. Yusupov and F. S. Hakimzyanov, with Norman Golb and Omeljan Pritsak, Khazarian Hebrew Documents of the Tenth Century (1982), and the Ibn Fadlan and Balkhi-school geographers in their standard editions as the outsider control.

Prediction clause (verbatim)

Prediction: a census of primary written sources for Khazaria and Volga Bulgaria before 1300 will find that texts composed by outsiders outweigh securely native-authored texts by more than ten to one in word count, and that every securely native written witness in the Bulgar language is epigraphic and funerary rather than documentary or literary (primary clause: the all-epigraphic character of native Bulgar witness; a native Bulgar documentary or literary text falsifies it).

Kill-dataset (verbatim)

Kill: the Volga Bulgar epitaph corpus in the editions of G. V. Yusupov and F. S. Hakimzyanov, with Norman Golb and Omeljan Pritsak, Khazarian Hebrew Documents of the Tenth Century (1982), and the Ibn Fadlan and Balkhi-school geographers in their standard editions as the outsider control.

Novelty / leakage triage

already answered in the literature

That the entire securely native written record of the Volga Bulgar language is the corpus of thirteenth-fourteenth-century Islamic gravestones - with no Bulgar documentary or literary text known - is a stated fact of the field, established in Erdal's edition of the epitaph language and the Yusupov/Hakimzyanov corpora. The primary clause (all native Bulgar witness epigraphic and funerary) is therefore already the published position, and the outsider-heavy character of the Khazar-Bulgar record (Ibn Fadlan, the Balkhi geographers, the contested Khazar Hebrew documents in Golb-Pritsak) is equally in print. The answer, not just the materials, is stated.

Predictions

No prediction registered yet.