The translator's ledger outlived both shelves

Claim (verbatim)

In 856, revising the count in his last years, Hunayn ibn Ishaq wrote Ali ibn Yahya a letter that is the pre-print world's finest census of a translation movement: the Risala, cataloguing 129 works of Galen one by one - whether he found Greek manuscripts and where, who had made a Syriac version and for which physician, who made the Arabic and for which patron (Bergstraesser's 1925 edition with the 1932 supplement on the second recension; John Lamoreaux's 2016 edition and translation). Run that ledger against the modern registry of what survives - Fichtner's Corpus Galenicum bibliography and Sezgin's GAS III - and the translation movement's harvest becomes countable. Byzantine copying followed the Alexandrian teaching canon, the sixteen-book curriculum and its satellites; Baghdad translated the whole census, curriculum or not. So the works that survive complete only in Arabic should be precisely the off-syllabus tail - the Hippocratic commentaries and occasional treatises no Greek classroom recopied: the commentary on Epidemics II, the commentary on Airs Waters Places, On Medical Experience, the compendium of Plato's Timaeus, On Examinations by Which the Best Physicians Are Recognized, On Medical Names, the later books of Anatomical Procedures. Where the syllabus ended, the only continuing custodian of Galen was a ninth-century Nestorian's workshop, and his own ledger is the instrument that proves it.

Prediction clause (verbatim)

Prediction: joining the Risala's 129 entries per work to Fichtner's Corpus Galenicum and Sezgin GAS III, at least 10 works of the census will survive complete in Arabic while no complete Greek text survives (primary clause: the >=10 Arabic-only-complete count; the verdict follows it), and at least half of that Arabic-only class will lie outside the Alexandrian sixteen-book curriculum as the Summaria Alexandrinorum roster defines it; a work counts Arabic-only only if the Greek lacks a complete continuous text (Greek fragments allowed), and the test voids for coverage if fewer than 100 of the 129 entries can be resolved to a survival status.

Kill-dataset (verbatim)

Kill: Hunayn's Risala (Bergstraesser 1925/1932; Lamoreaux 2016) joined work-by-work to Fichtner's Corpus Galenicum bibliography (BBAW) and Sezgin GAS III - a three-column Greek/Syriac/Arabic survival census with an Alexandrian-curriculum flag from the Summaria Alexandrinorum roster.

Novelty / leakage triage

already answered in the literature

Per-work survival of the Risala's 129 is already tabulated terrain: Bergstraesser's apparatus and Lamoreaux's notes annotate each entry's fate, Sezgin GAS III and Fichtner's Corpus Galenicum carry per-work survival notices, and the Arabic-only-complete class is an enumerated roster in the standard surveys - the commentaries on Epidemics II and on Airs Waters Places, On Medical Experience, the Timaeus compendium, On Examinations, On Medical Names, Anatomical Procedures IX-XV, On Character Traits, plus the lost-Greek tail of the Epidemics VI commentary already carry the list to the threshold, and the CMG's Supplementum Orientale series exists to edit precisely this class. The >=10 count is stated-or-guaranteed list-arithmetic on printed tabulations; only the Summaria Alexandrinorum curriculum flag is a join not yet run, and it is secondary.

• G. Bergstraesser, Hunain ibn Ishaq ueber die syrischen und arabischen Galen-Uebersetzungen (AKM 17.2, Leipzig, 1925), with Neue Materialien (AKM 19.2, 1932)
• J.C. Lamoreaux, Hunayn ibn Ishaq on His Galen Translations (Provo, 2016), annotations
• F. Sezgin, Geschichte des arabischen Schrifttums III (Leiden, 1970), Galen section
• G. Strohmaier, Der syrische und der arabische Galen, ANRW II.37.2 (1994)

Predictions

No prediction registered yet.