AI-generated conjecture · below the evidence/publication boundary

The census of missing clay

Status: Anticipated · untested

Status is derived only from the shepherd-authored triage/prediction data above -- community submissions and claims are a separate overlay and can never change it (see the participation panel below).

This is a proposed connection between two domains, generated by a language model. It is not an article and not evidence: it sits below the evidence/publication boundary. A quantitative prediction and a named kill-dataset are attached (when registered) so the claim stays falsifiable rather than merely evocative.

Claim (verbatim)

Ecologists count fish they cannot see by capture-recapture: tag a sample, resample, and the overlap reveals the population. Ur III bookkeeping recorded single transactions redundantly — a delivery could generate a receipt, a ledger entry, and a line in an annual account — and those redundancies are matchable by date, commodity, quantity, and personnel. The conjecture: capture-recapture across matched duplicates yields the first defensible estimate of how much documentation the Ur III state originally produced, and it will show that survival is savagely uneven by text type, differing by fivefold or more between genres. If it holds, the celebrated 100,000-tablet snapshot is revealed as a biased sample whose bias is finally measurable, and every aggregate statistic ever computed from Ur III data acquires an error bar.

Prediction clause (verbatim)

Cross-matching transactions between primary documents and summary accounts in the major Ur III archives will yield Lincoln-Petersen estimates of original documentation volume per genre. Primary clause, which decides the verdict: estimated survival rates differ across text genres by at least a factor of 5 with non-overlapping confidence intervals. Secondary clause: the overall estimated survival rate for the Drehem archive falls below 25 percent, implying an original output above 400,000 documents for its recorded decades.

Kill-dataset (verbatim)

BDTNS (transaction-level fields — date, commodity, quantity, personal names, text type — for the Drehem, Umma, and Girsu archives).

Nobody has run this test. The kill-data is named above. If you can run it — or you know the paper that already settles it — claim the kill or submit the prior. Kills and priors are credited here, by name, as they come in.

In the atlas

This conjecture is bridged, as an L1 lead, onto these Inferpedia subject pages.

Provenance

Run: Fresh agent generation · model: claude-fable-5

Composed blind by claude-fable-5 from internal knowledge only, with zero tool calls, and emitted directly as a single JSON text message.

Novelty / leakage triage

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

Searched capture-recapture/unseen-species estimation of lost documents. Kestemont et al. (Science 2022) applied ecological abundance estimators to manuscript survival of medieval literature, directly anticipating the method; no application of Lincoln-Petersen matching to redundant Ur III administrative documentation was located.

Kestemont et al., Forgotten books: The application of unseen species models to the survival of culture, Science 375 (2022) 765-769

Predictions

No prediction registered yet.

Weigh in

No community feedback yet.

Add your take

Posted immediately (spam is removed). Community feedback is never an adjudicated verdict and never changes this conjecture's triage label or status above.