Titles without texts

Claim (verbatim)

The Old Babylonian Sumerian literary catalogues are lists that record compositions one per line by their opening words - the Louvre catalogue (AO 5393), the Nippur catalogues (N1-N7, the Philadelphia lists), and the Ur catalogue (UET), first edited by S. N. Kramer as 'The Oldest Literary Catalogue' (BASOR 1942) and gathered in ETCSL's own catalogue section and in epsd2. Each incipit is a claim that an entire composition once existed and circulated. But a catalogue is a library's index, and the poems it indexes lived on individual tablets that had to be physically recopied to survive; most were not. So the index should overshoot the recovery - the incipit-lists name more compositions than the excavated, edited corpus has ever handed back. The loss here is not inferred from silence; it is written down as a list of names whose bodies are gone. The mechanism is the ordinary attrition of single-copy literary tablets against a catalogue that was, by design, complete for its own library at its own moment.

Prediction clause (verbatim)

Prediction: cross the incipits enumerated across the Old Babylonian literary catalogues against the composition inventory of the ETCSL/epsd2 Sumerian literary corpus; at least 25 percent of catalogued incipits will have zero identifiable recovered manuscript (primary clause: the unmatched-incipit fraction >= 0.25; the verdict follows it). Disambiguation: an incipit counts as recovered if it matches the first line of any edited composition in ETCSL or epsd2 under normalized-sign comparison, absorbing orthographic variants. Coverage guard: if fewer than four catalogues bearing >= 200 total incipit lines survive legibly, the test voids as a small-sample failure rather than resolving, since too few lists cannot separate real loss from catalogue idiosyncrasy.

Kill-dataset (verbatim)

Kill: the ETCSL catalogue section plus the epsd2 literary incipit index - normalized-sign match of catalogue incipits to edited-composition first lines, counting the unmatched. Runnable this week against these fixed published corpora.

Novelty / leakage triage

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

That the Old Babylonian catalogues name compositions never recovered is stated from Kramer's editio princeps onward, and the per-entry identification status is itself published - ETCSL's editions of the catalogues annotate each incipit with its identified composition where one exists, and Delnero's curricular study re-tabulates the major Nippur, Louvre, and Ur lists - but no located publication states the aggregate unmatched-incipit fraction across the catalogues against the ETCSL/epsd2 corpus, so the >= 0.25 arithmetic as posed is un-run in print.

• S.N. Kramer, 'The Oldest Literary Catalogue: A Sumerian List of Literary Compositions Compiled about 2000 B.C.', BASOR 88 (1942)
• P. Delnero, 'Sumerian Literary Catalogues and the Scribal Curriculum', ZA 100 (2010)
• J.A. Black et al., The Electronic Text Corpus of Sumerian Literature (Oxford, 1998-2006), catalogue section (c.0.2)

Predictions

No prediction registered yet.