The trough the empire fell into

Bronze Age Dated c. 2000 BCE – 1100 BCE Mesopotamia

Status: Already answered

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 conjecture imagined by a language model — drawn from its trained weights and held to falsifiability, novelty, and plausibility, not to any one method: it may join two or more fields, or none. 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)

Plot dated cuneiform by period and a canyon opens after the Old Babylonian collapse. Tablets thin dramatically across the First Sealand Dynasty and the early Kassite 'dark age' (roughly 1730-1450 BCE) before the Middle Babylonian recovery. This discontinuity can be production - a real political-economic collapse suppressing the making of documents - or survival - the relevant southern and Sealand sites are drowned, unexcavated, or looted. The histogram cannot by itself separate these, but its sharpness constrains the choice, and its sharpness is measurable in the catalogue right now. The mechanism at issue is precisely the ambiguity the coverage guard is built to adjudicate: a genuine ancient absence versus an accident of where the spade has fallen.

Prediction clause (verbatim)

Prediction (in-house CDLI copy, 126,000 rows): the dated-tablet count falls from the Old Babylonian bin (in-house: 14,975, plus 1,573 Early Old Babylonian) to the Middle Babylonian bin (in-house: 1,790) by a factor >= 5 (in-house actual ratio about 8; primary clause: Old-Babylonian:Middle-Babylonian dated-tablet ratio >= 5); and the trough is a genre-general collapse rather than one archive vanishing - the ratio remains >= 3 within the administrative genre alone. Disambiguation: 'Old Babylonian' and 'Middle Babylonian' are CDLI's own period labels; count distinct catalogued tablets, not fragments. Coverage guard: report alongside the number of distinct proveniences per period - if the Middle Babylonian site count is itself >= 5 times smaller, the drop is excavation rather than history, and the item resolves toward survival-accident rather than ancient absence.

Kill-dataset (verbatim)

Kill: the CDLI period field, dated-tablet histogram with per-period provenience counts. Runs THIS WEEK in-house, confirmable on the full CDLI catalogue.

Browse the registry: Cuneiform archives (CDLI) →

How this item is catalogued

Catalogue tags are curatorial metadata that make the conjecture findable. They are not findings, not a novelty verdict, and naming a source family is not a promise that the data exist or are accessible. Each dimension below says whether it was assigned, reviewed with none applying, uncertain, or not yet reviewed, and who recorded it.

Place & era tags — Assigned: Mesopotamia; Bronze Age Tagged by claude-fable-5
Dating — Phenomenon: Dated c. 2000 BCE – 1100 BCE The Old Babylonian collapse and Kassite 'dark age' trough, spanning the Old Babylonian period (c.2000-1600 BCE) through the Middle Babylonian/Kassite recovery (to c.1150 BCE). Tagged by claude-sonnet-5 + claude-fable-5 (gate)
Subject — Assigned: The cuneiform world (from the generation section)
Finer subject topic — Assigned: Cuneiform archive formation and taphonomy The object is a trough in the excavated tablet record and whether burial and digging explain it, squarely that page's subject. Tagged by claude-opus-5 draft (requested max effort) adjudicated by claude-fable-5-1 driver, lane 1da5; runtime not independently attested
Claim level — Assigned: Record claim The primary clause is a dated-tablet ratio between two periods, a claim about the density of the surviving ancient record, with the provenience guard able to resolve it toward modern excavation instead. Tagged by claude-opus-5 draft (requested max effort) adjudicated by claude-fable-5-1 driver, lane 1da5; runtime not independently attested
Method — Reviewed — none applies: An Old-Babylonian-to-Middle-Babylonian dated-tablet ratio of at least 5, holding at 3 within the administrative genre; the provenience counts are reported as a coverage guard to be read by eye, not as a bias term applied to the histogram. Tagged by claude-opus-5 draft (requested max effort) adjudicated by claude-fable-5-1 driver, lane 1da5; runtime not independently attested
Evidence source — Assigned: Cuneiform archives (CDLI) Kill test loads the CDLI period field for a dated-tablet histogram with per-period provenience counts. Tagged by claude-opus-5 draft (requested max effort) adjudicated by claude-fable-5-1 driver, lane 1da5; runtime not independently attested
Research stage — Idea — no test designed

On Inferpedia

This conjecture is linked to the following pages on Inferpedia, an encyclopedia of the missing — working atlas pages, some still early scaffolding.

Provenance

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

Fresh blind generation by claude-fable-5, 2026-07-17, cuneiform instrument-anchored wave against CDLI (including the in-house catalogue copy), ORACC, BDTNS, and Archibab: every Kill names a specific open corpus and a countable operation (catalogue-incipit vs recovered census, series colophon N-of-M completeness arithmetic, un-enrichment residue, directional letter-asymmetry, dated-histogram discontinuity, lexical-vs-documentary attestation gap, intake-vs-holdings, curricular copy-frequency, godlist-vs-cult, chronology-vs-attestation, forgery-by-genre, provenance-null-by-genre, letter tablet-census, and genre/period composition shares). Disjoint by construction from the 30 cuneiform conjectures of wave w22 (grep-checked per key name and per instrument-operation). In-house CDLI copy = 126,000 rows (about 30 percent of CDLI's 421,501), harvested 2026-07-04; exact in-house aggregates cited in the runnable-this-week items: genre-blank 10.4 percent, language-undetermined 14.3 percent, provenience-null 11.9 percent, flagged fakes 160 (0.127 percent), Administrative 70.9 percent, Ur III 47.4 percent, Old-Babylonian:Middle-Babylonian dated ratio about 8. Six items resolve THIS WEEK against the in-house copy (ordinals 3, 5, 11-partial, 12, 13, 15); two carry an honest 'Kill (not yet built)' where the cross-instrument join is real in sources but not yet a dataset (ordinals 6, 7, 14).

Novelty / leakage triage

already answered in the literature

Self-measurement by construction: the primary clause's ratio was run in-house during design - the item and the register both state the in-house OB:MB dated ratio of about 8 against a threshold pinned at >= 5. The field has also published the shape of the trough: Streck's quantification of the cuneiform corpus by period and Brinkman's catalogue of dated Kassite material document the post-Old-Babylonian documentary collapse. Only the within-administrative ratio and the per-period provenience guard remain un-run.

Sources cited by the triage

In-house CDLI catalogue copy (126,000 rows, about 30 percent of CDLI's 421,501 artifacts, harvested 2026-07-04), the wave's own registered aggregates: Old Babylonian 14,975 (+1,573 early OB) vs Middle Babylonian 1,790 dated rows
M.P. Streck, 'Grosses Fach Altorientalistik: Der Umfang des keilschriftlichen Textkorpus', MDOG 142 (2010)
J.A. Brinkman, Materials and Studies for Kassite History, vol. 1 (Chicago, 1976)

Predictions

No prediction registered yet.

Weigh in

No community feedback yet.

New here? Create an account first

Create an account or sign in and your feedback is tied to you — you can track it, get replies, and claim this conjecture so others know you’re working on it. Prefer not to? Just leave your take below as a guest — only the name you type is shown.

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.