The receipts weigh more than the library

Claim (verbatim)

Assurbanipal assembled the Nineveh libraries partly by confiscation. Administrative records and royal orders document the systematic collection of tablets and writing-boards from Babylonian scholars and temples: the acquisition and booty records, the scholarly correspondence ordering collection, and accession dockets that count incoming tablets and le'u writing-boards. Two materials entered that library - clay tablets, durable, and wax-on-wood writing-boards (le'u), almost wholly perishable. The accession record counts BOTH; the excavated library preserves essentially only the clay. So the library's own paperwork of intake should outweigh its surviving shelves, and the writing-boards it logged should be nearly invisible among recovered objects - a loss legible in the mismatch between a documented intake and a durable-only survival. The mechanism is a material-survival filter operating on a record that predates the filter: the intake list counted objects the ground could not keep.

Prediction clause (verbatim)

Prediction: the acquisition and administrative records' stated intake of tablets-plus-writing-boards exceeds the recovered Nineveh library tablet count by a factor >= 2, AND writing-boards constitute >= 10 percent of the documented intake yet < 1 percent of recovered library objects (primary clause: the writing-board recovery deficit - documented share >= 10 percent versus recovered share < 1 percent). Disambiguation: 'documented intake' comes from the accession and collection records' own counts; 'recovered' from the Nineveh library holdings. Coverage guard: accession records may double-count or round, so report intake as a range and treat the writing-board material-survival gap - not the raw total - as the primary, more robust clause.

Kill-dataset (verbatim)

Kill (not yet built): the SAA / SAAo acquisition and administrative texts and the Nineveh library holdings are both published, but no single dataset joins 'documented intake' to 'recovered holdings' - it must be compiled from the State Archives of Assyria volumes and the British Museum Nineveh (Kuyunjik) catalogue. Real sources; the join is not yet data.

Novelty / leakage triage

already answered in the literature

The primary clause is already in print: Parpola's publication of the Assyrian library records tabulates the documented intake - on the order of 2,000 tablets plus roughly 300 le'u writing-boards, a board share above 10 percent by his own published totals - and Frame and George's edition of Assurbanipal's tablet-collecting records reiterates that the logged boards are essentially absent among recovered Kuyunjik holdings. The material-survival mismatch is the stated point of both publications; only the formal intake-to-holdings join is unbuilt, and the separate intake-exceeds-holdings clause runs against the published counts (see report).

• S. Parpola, 'Assyrian Library Records', JNES 42 (1983)
• G. Frame and A.R. George, 'The Royal Libraries of Nineveh: New Evidence for King Ashurbanipal's Tablet Collecting', Iraq 67 (2005)

Predictions

No prediction registered yet.