The key outlived the doors

Claim (verbatim)

The tool that lets moderns read ancient Greek music at all is the Eisagoge of Alypius, a late-antique handbook transmitting complete correspondence tables between notation signs and pitches: paired vocal and instrumental sign-columns for each of the fifteen tonoi across the genera - well over a thousand sign-entries, printed in Jan's Musici Scriptores Graeci (1895), with lacunae toward the end. When early modern and modern scholars deciphered the Mesomedes hymns, the Delphic paeans, and the Seikilos stone, Alypius was the key. The inversion is total: the decoder survived by continuous copying - a compact technical reference inside the theory curriculum, exactly what scriptoria kept - while the encoded archive it was written to serve, the actual written music of a millennium, almost entirely vanished. That leaves a measurable quantity: unused key-space. Alypius defines sign-tables for fifteen tonoi, but the surviving corpus is so thin that most tables now decode nothing that still exists - whole tonoi with complete transmitted sign-columns and zero surviving documents that use them. The mechanism: reference works survive in proportion to curricular utility, documents in proportion to reuse, and the two rates are independent, so a reference can outlive its entire object-domain. We are holding a full cipher key to a burned archive - able to read fluently, with almost nothing left to read - and the fraction of the key that still opens anything is a direct measurement of how much of the archive burned.

Prediction clause (verbatim)

Prediction: tabulating the notation-key (tonos) assignments given in Pöhlmann-West's per-item commentary against Alypius' fifteen tables, at most eight of the fifteen tonoi will be attested by even one surviving item, so at least seven complete transmitted sign-tables decode zero surviving documents (primary clause: the <=8-of-15 attested-tonoi bound; the verdict follows it), and the two most frequent keys (expected: Lydian and Hypolydian) will jointly account for more than 40 percent of key-assignable items; assignments follow the edition's commentary, with items of indeterminable key excluded from the denominator; the test voids if fewer than 30 items receive key assignments.

Kill-dataset (verbatim)

Kill: Alypius' tables in Jan's Musici Scriptores Graeci (1895) crossed with the per-item key discussions in Pöhlmann-West's Documents of Ancient Greek Music - a tonos-attestation tabulation over the whole corpus.

Novelty / leakage triage

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

The decoder-without-documents inversion is stated - Jan prints Alypius' fifteen-tonoi sign tables, and West's handbook notes both the tables' survival and the clustering of the extant documents in a few keys, with Lydian and Hypolydian predominant - but the specific instrument here, a tonos-attestation tabulation of all fifteen Alypian tables against the edition's per-item key assignments with a <=8-of-15 attested bound and a >40 percent two-key share, is exactly the un-run attestation count; no publication was located that scores empty versus attested tables across the whole corpus.

• C. von Jan, Musici Scriptores Graeci (Leipzig, 1895), the Alypius tables
• M.L. West, Ancient Greek Music (Oxford, 1992), on the keys of the surviving documents
• E. Pohlmann & M.L. West, Documents of Ancient Greek Music (Oxford, 2001), per-item key discussions

Predictions

No prediction registered yet.