The digest buried its originals

Claim (verbatim)

The Excerpta were commissioned to make the historians useful; the effect was to make them redundant. Once the palace held the essential matter of a thirty-book history sorted under subject heads, the incentive to recopy the unwieldy original - a shelf of quires with no classroom constituency - fell to nothing, and the full texts of the late antique classicizing historians drop out of the record within about a century of the excerpting. The proof lives in the fragment arithmetic of the modern editions: Priscus of Panium, Malchus, Menander Protector, Peter the Patrician, Nicolaus of Damascus and Eunapius survive overwhelmingly as Constantinian excerpts - R. C. Blockley's Fragmentary Classicising Historians of the Later Roman Empire (2 vols, 1981-1983) and his History of Menander the Guardsman (1985), Thomas Banchich's Lost History of Peter the Patrician (2015), and FGrH 90 for Nicolaus are, in large part, editions of the Excerpta rearranged by author, with the Suda's contributions themselves Excerpta-derived (de Boor). The digest is simultaneously the killer and the grave-good: without it we would have almost nothing of these authors; because of it, nothing else of them was kept. Dexippus is the exception that dates the process - his Scythica resurfaced in a Vienna palimpsest undertext (the Scythica Vindobonensia), the one kind of carrier that escapes the digest's economy.

Prediction clause (verbatim)

Prediction: for the pinned six - Eunapius' Histories, Priscus, Malchus, Menander Protector, Peter the Patrician, and Nicolaus of Damascus' Historiae - at least five will have more than half of their edited Greek fragments, by fragment count, transmitted solely through the four surviving Excerpta collections (primary clause: at least 5 of 6 authors above the 0.5 Excerpta-carrier share; the verdict follows it); secondary clause: none of the six has any direct-tradition witness in Pinakes; disambiguation: fragment counts follow Blockley's, Banchich's and FGrH numeration; testimonia are excluded; Suda-derived fragments are counted separately and not as Excerpta, so the primary is conservative; coverage guard: void for any author whose standard edition cannot be partitioned by carrier.

Kill-dataset (verbatim)

Kill: the fragment-source registers of Blockley I-II, Blockley's Menander, Banchich's Peter the Patrician and FGrH 90, joined to Pinakes for the zero-direct-witness check - per-author carrier-attribution counts.

Novelty / leakage triage

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

That these six survive overwhelmingly as Constantinian excerpts is stated in the editions themselves, and the digest-buried-the-originals thesis is old (Lemerle's encyclopedism chapter; Nemeth weighs it critically); but the pinned arithmetic is open: with Suda-derived fragments deliberately not counted as Excerpta, the strictly-Excerpta fragment share for the Suda-heavy Eunapius and the Athenaeus-and-Josephus-fed Nicolaus is nowhere computed, and five-of-six above 0.5 could honestly fail. The count over Blockley's, Banchich's and Jacoby's source registers is new, and the zero-direct-witness secondary is the only guaranteed part.

• R.C. Blockley, The Fragmentary Classicising Historians of the Later Roman Empire, 2 vols (Liverpool, 1981-1983), introductions and source registers
• T.M. Banchich, The Lost History of Peter the Patrician (London, 2015)
• A. Nemeth, The Excerpta Constantiniana and the Byzantine Appropriation of the Past (Cambridge, 2018), on the fate of the excerpted originals

Predictions

No prediction registered yet.