Phonograms funnel

Claim (verbatim)

Phonograms funnel. The man'yogana graph-per-syllable inventory should narrow as poetry recording became clerical. Early Man'yoshu strata record poems through whichever graphs a given scribe's Chinese training suggested — a wide, idiosyncratic inventory — while later compilation strata are quasi-chancery work, and every chancery converges its script. The funnel should track the compilation stratum, not the poem's composition date: it is a fact about scribes, not poets, and it should appear only in the phonographic channel, not in logographic writing, which had no equivalent standardization pressure.

Prediction clause (verbatim)

Group Man'yoshu poems by the standard chronological strata and book-groupings; compute per-stratum entropy of graph choice per syllable over phonographically written poems. Primary clause (verdict follows it): entropy declines monotonically across the four standard strata, with a total decline of at least 30 percent. Secondary: logographic (borrowed-reading) writing within the same strata shows no comparable decline, isolating the funnel to the phonographic channel.

Kill-dataset (verbatim)

Kill: the Oxford-NINJAL Corpus of Old Japanese (ONCOJ) and NINJAL's digitized Man'yoshu with graph-level annotation — open and queryable without new ingestion.

Novelty / leakage triage

provisional — model-triaged, shepherd review pending

A provisional first pass authored by the model (Opus), not yet confirmed by the shepherd. It carries the same dated-search requirements as an authoritative verdict but is excluded from every headline figure and cannot underwrite a prediction until a shepherd confirms it. Provisional reading: Adjacent (closely related prior work exists).

The premise is documented: early Man'yoshu phonographic writing used a wide, idiosyncratic man'yogana inventory (about 970 graphs for 90 morae) chosen on the writer's whim with no standard, and the long-run trajectory of Japanese phonographic writing is toward a reduced, standardized set (eventually the kana). Not located: the item's specific stratum-level measurement isolating the mechanism -- per-syllable graph-choice entropy declining MONOTONICALLY across the four standard chronological strata by >=30% total, tracking the compilation stratum (scribes) rather than composition date (poets), and appearing ONLY in the phonographic channel while logographic writing shows no comparable decline. The narrowing trajectory is recognized; the stratum-conditioned, channel-isolated entropy measurement over ONCOJ is unrun.

• Man'yogana overview -- ~970 graphs for 90 morae, no standard, choice on the writer's whim; the trajectory toward a reduced phonographic set — The wide early inventory and the narrowing trend -- the established datum; the stratum-level channel-isolated entropy measurement is unrun
• Oxford-NINJAL Corpus of Old Japanese (ONCOJ) and NINJAL's graph-annotated Man'yoshu — The kill dataset for the per-stratum graph-choice entropy; open and queryable

Predictions

No prediction registered yet.