AI-generated conjecture · a wager, not a finding
← All conjectures · Philosophy, theology & the schools
Condemned in Greek, canonized in Syriac
Status is derived only from the shepherd-authored triage/prediction data above -- community submissions and claims are a separate overlay and can never change it (see the participation panel below).
Claim (verbatim)
For a Greek theologian, formal condemnation decided not whether his works survived but in WHICH LANGUAGE. The empire could burn books inside its jurisdiction - Theodosius II ordered Nestorius' writings destroyed in 435, Justinian criminalized possessing Severus of Antioch in 536, the Three Chapters council of 553 anathematized Theodore of Mopsuestia and the teaching of Diodore of Tarsus - but the Syriac and Armenian churches beyond Chalcedonian enforcement went right on copying their heroes in translation. Translation was extraterritorial custody. The Clavis Patrum Graecorum (Geerard and Noret, five volumes with Supplementum) makes this countable, because it registers per work the languages of survival, marking the stratum that survives only in versions - Syriac, Armenian, Coptic, Georgian, Arabic - with no Greek text. Nestorius' apologia lives in Syriac; Theodore's catechetical homilies and biblical commentary chiefly in Syriac; Timothy Aelurus' refutation of Chalcedon in Armenian and Syriac; Severus' vast corpus almost wholly in Syriac; while John Chrysostom, Basil, Gregory of Nazianzus and Cyril - translated far MORE, since orthodox bestsellers travelled too - kept their Greek as well. So the versio-only share of a corpus should split cleanly by condemnation status: for the condemned it is the majority mode of survival; for the orthodox it is a thin residue. The operation runs on the repertorium's own language notices - the translation-side census of what anathema did.
Prediction clause (verbatim)
Prediction: over qualifying CPG entries (works with a complete text surviving in some language), the pooled versio-only share of the condemned roster - Nestorius, Theodore of Mopsuestia, Diodore of Tarsus, Severus of Antioch, Timothy Aelurus - will exceed 60% and be at least four times the pooled share of the orthodox controls - John Chrysostom, Cyril of Alexandria, Basil of Caesarea, Gregory of Nazianzus - which will sit below 15% (primary clause: the pooled >=4x separation with the two absolute bounds; the verdict follows it); secondarily, every individual author clearing 5 qualifying entries will fall on his roster's side of 40%; a work is versio-only when no complete Greek text survives but a complete version does; texts surviving in Greek only inside the hostile dossier - conciliar acta and quoting refutations - are excluded from both numerator and denominator, since transmission-in-refutation is the enemy's channel, not copying demand; dubia and spuria are excluded per the CPG's own classifications, works composed in a language other than Greek are excluded, and the test voids for coverage if the condemned pool resolves to fewer than 25 qualifying entries or the control pool to fewer than 40.
Kill-dataset (verbatim)
Kill: the Clavis Patrum Graecorum with Supplementum, reading each work's language notices - a per-author versio-only share census across the two named rosters.
Nobody has run this test. The kill-data is named above. If you can run it — or you know the paper that already settles it — claim the kill or submit the prior scholarship. Kills and prior scholarship are credited here, by name, as they come in.
Provenance
Run: Fresh agent generation · model: claude-fable-5
Fresh blind generation by claude-fable-5, 2026-07-18, translation-as-survival-channel wave: every kill names a census instrument of the translation traffic (Hunayn's Risala, CPG language notices, critical-edition transmission registers, fragment-collection carrier apparatus) and a countable operation, thresholds far from 1 with coverage guards. Discipline and drops: no Sanskrit-into-Chinese/Tibetan item at all - the catalogue ground is owned (Kaiyuan-lu queben by breadth-india ord 3, An Shigao attribution by eastasia-ctext ord 6, Tanjur counting by breadth-india-w2 ords 2/13 and breadth-seasia-w2 ord 1); a Nagarjuna two-canon disjointness candidate was dropped for sharing those instruments. The condemned-authors item is carved to the TRANSLATION side (CPG versio-only shares) - Greek-side survival of the condemned is owned by byzantine-dbbe-pinakes ord 3 (Photios x Pinakes/TLG); the within-work book-cliff item is carved off dbbe-pinakes ord 8 (Greek historians' block-contiguity via Pinakes) onto dual-stream mathematics/medicine with a commentary-boundary mechanism and edition-prolegomena instruments. No Syriac-Organon curricular operation (w02 ord 27 owns vHMML logic-block copying), no Pinakes-x-vHMML bestseller-export join (w10 ord 2), Evagrius kept out of the condemned roster (w10 ord 21 owns his attribution jurisdiction); Sicilian translation codicology untouched (w20 ords 7/24 own the PAL convoy/parasite operations) - the Sicilian item here is internal table arithmetic in Ptolemy's Optics; no Judeo-Arabic-to-Hebrew geography (w02 ord 6); Fihrist used only as an instrument (hapax mortality owned by w03 ord 10, reuse afterlife by islamicate-openiti ord 1). Hunayn's Risala anchors three items under three disjoint operations (three-column survival census; Syriac-layer die-off; scarcity-notice double-death predictor), flagged as deliberate. Item 17 extends the language set beyond the steered five channels (Aramaic-Greek-Geez chain) - grep-clean ground. Honest confidence flags: the Pappus-coverage fraction (ord 5), the Nestorius fragment-concordance count (ord 13) and the Abydenus carrier share (ord 15) are MODERATE and carried by guards or floor clauses; famous loss-facts are pinned to new arithmetic throughout.
Novelty / leakage triage
anticipated in the literature — this exact test has never been run
The direction is textbook - condemned authors surviving in Syriac and Armenian versions while orthodox bestsellers kept their Greek is the standard narrative of Baumstark and every patrology - and the CPG's language notices make it countable; but no located study pools versio-only shares per author over CPG entries and tests a condemned-versus-control separation with pinned thresholds (>60%, <15%, >=4x). Geerard's introductions do not tabulate shares, and no CPG-based quantitative survey of this kind was found; the pooled cross-tabulation is genuinely un-run. Two roster wrinkles for resolution: Diodore's principal complete survivor (the Psalms commentary) is transmitted in Greek, and Theodore's Twelve Prophets commentary survives complete in Greek - the pooled clause can absorb both, but the per-author sub-clause leans on the under-5-entry exemption.
- M. Geerard, Clavis Patrum Graecorum, 5 vols (Turnhout, 1974-1987), with M. Geerard & J. Noret, Supplementum (Turnhout, 1998)
- A. Baumstark, Geschichte der syrischen Literatur (Bonn, 1922)
- J.-M. Olivier, Diodori Tarsensis Commentarii in Psalmos I (CCSG 6, Turnhout, 1980)
Predictions
No prediction registered yet.
Weigh in
No community feedback yet.
New here? Create an account first
Create an account or sign in and your feedback is tied to you — you can track it, get replies, and claim this conjecture so others know you’re working on it. Prefer not to? Just leave your take below as a guest — only the name you type is shown.
Add your take
Posted immediately (spam is removed). Community feedback is never an adjudicated verdict and never changes this conjecture's triage label or status above.