The sack cut deeper than the fall

Claim (verbatim)

Byzantine book production suffered two catastrophes with names, 1204 and 1453, and intuition ranks the extinction above the amputation. The dated-manuscript series should show the opposite. In 1204 the Fourth Crusade decapitated the production system itself - the capital's patrons, scriptoria, libraries and market - and the successor states spent a generation rebuilding: the Nicaean book famine that Vatatzes' and Blemmydes' documented library-hunting journeys tried to remedy. By 1453 the system had already emigrated: Cretan ateliers, the Italian humanist market and the emigre copyists around Bessarion were paying for Greek books at rates Constantinople could no longer match, and the fall relocated production it did not destroy - in dated Greek manuscripts, the later fifteenth century is a boom, not a grave. The two notches are countable in the corpora that have anchored Greek palaeography since Lake - Kirsopp and Silva Lake's Dated Greek Minuscule Manuscripts to the Year 1200 (1934-1945) and Alexander Turyn's dated-manuscript volumes for the Vatican, Italy and Great Britain (1964, 1972, 1980) - and, at far larger sample size, in the dated witnesses recorded in Pinakes. The comparison is a two-point calibration of what kills a manuscript culture: losing the capital hurt more than losing the empire, because the thing that copies books is a market with patrons, not a state with borders.

Prediction clause (verbatim)

Prediction: in the dated Greek manuscript series (Pinakes dated witnesses, cross-checked against the Lake and Turyn corpora), the mean per-decade count for 1204-1233 will fall at least 40% below the 1174-1203 mean, while the 1454-1483 mean will fall no more than 20% below the 1424-1453 mean - the proportional 1204 notch at least twice the 1453 notch (primary clause: the >=2x notch-depth ratio with the 1204 notch at >=40%; the verdict follows it); disambiguation: only manuscripts dated or datable to within a decade count; proportional drops are used so the two eras' different base rates cannot bias the comparison; provenance is unrestricted - relocated production counts as production; coverage guard: void if either comparison window holds fewer than 50 dated manuscripts.

Kill-dataset (verbatim)

Kill: Pinakes dated-witness extracts together with the Lake and Turyn dated-manuscript corpora - per-decade production histograms with a two-notch proportional comparison.

Novelty / leakage triage

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

Each notch is documented on its own - the post-1204 production collapse and the Nicaean book famine with Blemmydes' hunting journeys (Constantinides; Prato's palaeographic studies of Latin-period production), and the post-1453 relocation of copying to Crete and Italy (Geanakoplos and the emigre-scribe literature) - but the two-notch proportional comparison on a single dated series, with a pinned 2x depth ratio, is un-run; the Lake and Turyn corpora plus Pinakes dated witnesses make it computable for the first time at sample size.

• C.N. Constantinides, Higher Education in Byzantium in the Thirteenth and Early Fourteenth Centuries (Nicosia, 1982)
• D.J. Geanakoplos, Greek Scholars in Venice (Cambridge, MA, 1962)
• G. Prato, Studi di paleografia greca (Spoleto, 1994), studies of thirteenth-century Greek book production

Predictions

No prediction registered yet.