The copyist kept the words and dropped the tune

Claim (verbatim)

Greek lyric poetry is music - melos means song, and Sappho, Pindar, and Alcaeus composed melodies inseparable from their words. The papyrus record of these poets is substantial: the Leuven Database of Ancient Books registers on the order of two hundred witnesses of the canonical melic poets - the great Sappho finds, the Pindar rolls of Oxyrhynchus, Bacchylides' long London roll, Timotheus' Persae on one of the very oldest Greek book-papyri (P.Berol. 9875, fourth century BCE). Notation existed the whole time: by the mid-fourth century Aristoxenus is already polemicizing against people who equate notation-literacy with harmonic science, and third-century-BCE papyri of Euripidean drama carry notes. Yet the number of melic-poet papyri with musical notation is zero. Not one Sappho, Pindar, Alcaeus, Bacchylides, Anacreon, Simonides, or Timotheus witness carries a note: the book trade copied songbooks as text from the very first copies we possess. The mechanism: the book roll was a reading technology sold to a reading public, and melody was stripped at the point of commercial reproduction because buyers were readers, not performers; performance transmission stayed oral and professional and died on its own schedule. The written channel never carried classical lyric's music at all - which relocates the loss event: it did not happen in late antiquity's disasters but at the first copy, centuries earlier, as a market decision repeated every time a songbook was sold.

Prediction clause (verbatim)

Prediction: querying the Leuven Database of Ancient Books for papyrus and parchment witnesses of the nine canonical melic poets plus Timotheus, the witness count will exceed 150 while the subset bearing musical notation will be exactly zero (primary clause: the 0-notated-of->150-witnesses share; the verdict follows it), even though the same database's dramatic-text records include at least three notated Euripides witnesses proving the channel could carry notes; a witness counts as notated only if DAGM or its supplements accept its signs as musical notation; the test voids if the melic witness count falls below 100.

Kill-dataset (verbatim)

Kill: the Leuven Database of Ancient Books (trismegistos.org/ldab) author queries for the melic poets, crossed against the notated roster in Pöhlmann-West's Documents of Ancient Greek Music - a set intersection whose predicted size is zero.

Novelty / leakage triage

already answered in the literature

That the canonical melic poets came down as text without a note of their music is stated in the handbook literature: West discusses the total loss of the lyric classics' melodies and the fact that the earliest notated literary papyri are Euripidean, and the Pohlmann-West roster - the published tabulation of all accepted notation - contains no Sappho, Pindar, Alcaeus, Bacchylides, or Timotheus witness, which guarantees the zero-notated subset. The >150 melic witness count is the Leuven database's own author tally, and the verdict is pinned to the zero, so the un-run part is clerical.

• M.L. West, Ancient Greek Music (Oxford, 1992), on the loss of the lyric poets' music and the notated Euripides papyri
• E. Pohlmann & M.L. West, Documents of Ancient Greek Music (Oxford, 2001), roster containing no melic-poet item

Predictions

No prediction registered yet.