What the fire spared is what hands could carry

Claim (verbatim)

Ethiopia has made parchment books since late antiquity — the Garima Gospels, radiocarbon-dated to roughly the fifth to seventh centuries, prove how early — yet the national manuscript stock is overwhelmingly recent. The wall is the war of Ahmad b. Ibrahim al-Ghazi, 1529-43, whose own chronicle, the Futuh al-Habasha, itemizes great churches burned with their treasuries of books. What survived eleven centuries of production is what monks could carry at a run, and what a monastic community carries first is its Gospel book — the liturgically indispensable, most venerated class. The age profile of the surviving corpus should therefore show catastrophe rather than decay: a near-empty early millennium, then Gondarine abundance; and the thin pre-conquest stratum should be generically skewed toward Four Gospels codices far beyond that genre's share of later production. Prediction: in the dated and datable manuscript records of the Beta masaheft research environment, including the Ethio-SPaRe descriptions from Tigray — the oldest Christian landscape and the best case for early survival — manuscripts assigned before 1540 will be under 5% of holdings while post-1630 manuscripts outnumber them at least eightfold, and Four Gospels books will form at least a third of the pre-1540 stratum against under a tenth of the post-1630 one (primary clause: the under-5% pre-1540 share; the verdict follows it). Kill: the open TEI catalogue data of the Beta masaheft research environment (betamasaheft.eu), incorporating the Ethio-SPaRe project descriptions, controlled against the EMML microfilm catalogues (Getatchew Haile et al.); the computation is a date-bucket histogram with a genre cross-tabulation, with the Futuh al-Habasha (tr. Stenhouse, 2003) as the testimony of mechanism.

Prediction clause (verbatim)

Prediction: in the dated and datable manuscript records of the Beta masaheft research environment, including the Ethio-SPaRe descriptions from Tigray — the oldest Christian landscape and the best case for early survival — manuscripts assigned before 1540 will be under 5% of holdings while post-1630 manuscripts outnumber them at least eightfold, and Four Gospels books will form at least a third of the pre-1540 stratum against under a tenth of the post-1630 one (primary clause: the under-5% pre-1540 share; the verdict follows it).

Kill-dataset (verbatim)

Kill: the open TEI catalogue data of the Beta masaheft research environment (betamasaheft.eu), incorporating the Ethio-SPaRe project descriptions, controlled against the EMML microfilm catalogues (Getatchew Haile et al.); the computation is a date-bucket histogram with a genre cross-tabulation, with the Futuh al-Habasha (tr. Stenhouse, 2003) as the testimony of mechanism.

Novelty / leakage triage

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

The near-emptiness of the pre-Gragn stratum (attributed to the 1529-43 wars) and the seniority of Four Gospels books as the oldest surviving class (Garima) are standard knowledge, and the Ethio-SPaRe/Beta masaheft descriptions exist as open data; but no published date-bucket histogram with a genre cross-tabulation at anything like these thresholds was located.

• D. Nosnitsin, Churches and Monasteries of Tegray: A Survey of Manuscript Collections (Harrassowitz, 2013)
• J.S. McKenzie & F. Watson, The Garima Gospels: Early Illuminated Gospel Books from Ethiopia (Manar al-Athar, 2016)

Predictions

No prediction registered yet.