AI-generated conjecture · below the evidence/publication boundary

Last learned, first lost

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)

Sumerian stopped being spoken around 2000 BCE yet remained written for nearly two more millennia — but it retreated from writing genre by genre, not all at once. Second-language pedagogy supplies the mechanism: material drilled earliest and hardest is overlearned and cheapest to retain for life. The conjecture: written genres abandoned Sumerian in reverse order of curricular acquisition — the advanced literary Sumerian learned last went first, while the elementary layer of accounting terms, legal formulae, and list vocabulary drilled into every beginner persisted as frozen Sumerograms essentially forever. If it holds, the strange sandwich of later Babylonian writing — Akkadian prose stuffed with Sumerian word-signs — is a fossil of lesson order, and the timetable of a dead language's written afterlife becomes predictable from the syllabus of its schools.

Prediction clause (verbatim)

Ranking written genres by the date at which syllabically written Akkadian first exceeds 50 percent of content words within them, the resulting order will correlate with reverse curricular position (elementary-taught material crossing latest) with Spearman rho of at least 0.6 across at least six genre categories. Primary clause, which decides the verdict: the rank correlation is positive and significant at p<0.05. Secondary clause: legal operative formulae retain majority-Sumerographic writing at least 300 years longer than narrative literary prose.

Kill-dataset (verbatim)

CDLI and ORACC (period-tagged transliterated corpora permitting sign-by-sign Sumerographic versus syllabic tallies by genre from Ur III through Middle Babylonian).

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 Sumerogram retention by genre and period. Genre- and period-dependent Sumerogram density and the conservatism of scribal training are documented, but the specific claim that genres abandoned Sumerian in reverse curricular-acquisition order, with a quantitative crossover ranking, was not located.

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.