The millions shrink when the titles are counted twice

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)

The standard awe-statistic — millions of South Asian manuscripts, most catalogued only by title — quietly treats titles as works, but title-level cataloguing errs in both directions: one work travels under several titles (splitting), and one title covers several works (lumping). The two error rates are not symmetric in a tradition that abridged, honorifically retitled, and regionally renamed constantly: splitting should dominate lumping, so the true count of distinct works sits substantially below the title count. The bias is measurable wherever a title-level census overlaps a work-identity database, and its size converts the mythic millions into a defensible estimate. If it holds, the subcontinental census gap is smaller in works even though no smaller in objects — a correction that changes the loss-rate arithmetic for the entire field.

Prediction clause (verbatim)

Matching NGMPP/NGMCP title-level records against the work-identity graph of the Pandit database, the mean number of distinct catalogue titles per work will exceed the mean number of distinct works sharing an identical title by at least a factor of 2, implying a distinct-work count at least 25% below the raw title count for the matched population. Primary clause: the at-least-twofold asymmetry of titles-per-work over works-per-title; the 25% shrinkage figure is secondary.

Kill-dataset (verbatim)

Title-to-work reconciliation between the NGMPP/NGMCP catalogue and the work-identity records of the Pandit database.

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

Generated blind in a single Write with no reads, web access, or database queries; this is the second attempt for wave W14 after a prior instance died to a network error before writing its packet.

Novelty / leakage triage

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

The NCC's title-level cataloguing and the ~5-million manuscript awe-statistic are known, and title/work splitting-vs-lumping is a general cataloguing problem, but measuring the asymmetry (titles-per-work >=2x works-per-title; ~25% shrinkage in distinct-work count) via NGMPP-Pandit reconciliation is un-run. Thin field.

'Catalogus Catalogorum' (Wikipedia) — NCC as alphabetical register of Sanskrit works; ~5M mss estimate

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.