Ars Inquirendi

AI-generated conjecture · below the evidence/publication boundary

← All conjectures · Authors' minds & methods

The mill grinds its own flour

Status: Falsified

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)

Ibn al-Jawzi counted output the way a pietist counts prostrations — pages against days, pen-shavings hoarded to heat the water for washing his corpse. The interiority claim: for him, compilation was worship with a ledger, and no one maximizing sanctified page-count composes each page fresh; the method must be recombination, his own earlier books cut and reset. His corpus should behave like a mill grinding its own flour, at a self-reuse density ordinary prolific authors never approach. Prediction: in the KITAB text-reuse data over OpenITI, the within-author reuse density among Ibn al-Jawzi's distinct works (aligned-passage instances normalized by pairwise word volume) is at least three times the median within-author density of all OpenITI authors with twenty or more distinct works; primary clause: the threefold-versus-median comparison. Kill: the OpenITI corpus and the KITAB project's openly released pairwise text-reuse datasets on GitHub; the exact computation is the normalized within-author reuse density for Ibn al-Jawzi against the distribution over all authors with at least twenty works.

Prediction clause (verbatim)

Prediction: in the KITAB text-reuse data over OpenITI, the within-author reuse density among Ibn al-Jawzi's distinct works (aligned-passage instances normalized by pairwise word volume) is at least three times the median within-author density of all OpenITI authors with twenty or more distinct works; primary clause: the threefold-versus-median comparison.

Kill-dataset (verbatim)

Kill: the OpenITI corpus and the KITAB project's openly released pairwise text-reuse datasets on GitHub; the exact computation is the normalized within-author reuse density for Ibn al-Jawzi against the distribution over all authors with at least twenty works.

Provenance

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

Fresh blind generation instance (prompt file and existing-titles list only; no repository, web, or prior-art access), 2026-07-16, campaign Minds & Works wave M01.

Novelty / leakage triage

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

Ibn al-Jawzi's extreme productivity and reputation for self-recombination are known, and the KITAB project studies Arabic text-reuse, but the specific claim that his within-author reuse density is >=3x the median of authors with 20+ works was not located as a measured result.

  • S. Savant and the KITAB project, text-reuse studies (kitab-project.org)

Predictions

Killed registered 2026-07-16 calibration prediction (parent triage: leaked/adjacent)

Resolution: Killed

Caveats: ROBUSTLY KILLED. Ibn al-Jawzi's within-author self-reuse density sits at the MEDIAN of the 53 most prolific OpenITI authors (rank 24 of 53), not the >=3x outlier the conjecture staked -- he recombines his own books at about the rate his prolific peers do, not at a density 'ordinary prolific authors never approach'. The character-based density proxy (used to avoid re-fetching ~1,900 texts) was validated against an exact word-based computation for Ibn al-Jawzi himself: 1.10 vs 0.85, both an order of magnitude below the 3x threshold, so the approximation is immaterial. The one real judgement call -- constructing the >=20-works population (excluding grab-bag catalogue codes and translator-buckets whose 'works' are unrelated foreign originals) -- does not move the verdict (a classical-era-only median gives 0.89, still killed). The pietist-ledger picture is not borne out in the reuse data.

Registered against the KITAB project's openly released pairwise text-reuse data over OpenITI (kitab-project.org; github.com/OpenITI). Claim under test (primary clause): Ibn al-Jawzi's within-author self-reuse density is AT LEAST THREE TIMES the median within-author density of all OpenITI authors with >= 20 distinct works.

Resolution criteria: SHARED CONVENTIONS (identical across all rows of this packet). (a) TEXT SELECTION: use the PRIMARY OpenITI version of each named work (the '-ara1' primary per OpenITI convention, or the single most-complete version if no primary is flagged); if a work is split across several OpenITI version files, concatenate all parts of that same work's primary version in order. The build must FREEZE the exact OpenITI version URI(s) + a sha256 per work in a freeze_manifest before any metric is computed. (b) CLEANING: strip OpenITI mARkdown structural material (header block; lines/markers beginning with '#', '~~', page markers like 'PageVNN_NNN', 'ms', milestone markers 'Milestone###', editorial '@'-tags) to leave the running body text; strip Arabic tashkil (U+064B-U+0652, U+0670, U+0640 tatweel). (c) WORD = one whitespace-delimited token of the cleaned body text. (d) MARKER MATCHING is done on the tashkil-stripped text; alif/hamza variants are normalized for marker matching only (alif-forms أ إ آ ا -> ا) NOT for word counting. (e) All raw counts, rates, and the cleaned word totals per work must be reported alongside the verdict. (f) computed_at must postdate registered_at. || POPULATION: all OpenITI authors carrying >= 20 distinct works for which KITAB pairwise reuse alignments exist. WITHIN-AUTHOR DENSITY D(author) = (total aligned-passage words across all unordered within-author work-pairs) / (sum over those pairs of the word-length of the SMALLER work in the pair). [This normalizes shared material by the maximum that COULD be shared.] PRIMARY STATISTIC = D(Ibn al-Jawzi) vs MEDIAN of D over all qualifying authors. CLAUSE PRECEDENCE: (1) INCONCLUSIVE_BY_DESIGN if Ibn al-Jawzi has < 20 distinct works in the KITAB/OpenITI data, OR fewer than 15 OTHER authors qualify (>=20 works) for a stable median, OR KITAB reuse coverage of Ibn al-Jawzi's works is absent. (2) SUPPORTED if D(Ibn al-Jawzi) >= 3 x median(D). (3) KILLED if < 3x. Report D(Ibn al-Jawzi), the median, the qualifying-author count, and Ibn al-Jawzi's rank within the distribution.

Known priors disclosure: Held: the conjecture's reasoning (Ibn al-Jawzi maximized sanctified page-count by recombining his own earlier books) and general knowledge that KITAB publishes pairwise passim reuse over OpenITI and that Ibn al-Jawzi is among the most prolific OpenITI authors. NOT computed: his self-reuse density, the cross-author median, or his rank. Threefold-versus-median threshold is the conjecture's own.

KITAB one_to_all pairwise text-reuse data (kitab-project-org, versioned 2025.1.9) over OpenITI, pinned to the same commit as the Tier-1 build; per-passage source/target character offsets anchored to OpenITI's native ~300-word msNNNN milestone tags. Sonnet computed (symmetrized-dataset dedup + milestone-offset resolution independently validated in-build via 12 coherent samples). Fable independently re-derived the Ihya' quarter distribution from scratch (own milestone->quarter mapping) as a pipeline check. computed_at postdates registered_at (rule 6). Primary D is character-based (OpenITI char_length fields, no extra fetch); validated against an exact word-based D for Ibn al-Jawzi himself.

Dataset: KITAB within-author reuse density D for every OpenITI author with >=20 works (53 qualifying authors after excluding blank-author/grab-bag catalogue codes), normalized over ALL C(k,2) work-pairs (the registration's 'maximum that could be shared'). D(Ibn al-Jawzi) vs the population median.

computed 2026-07-17

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.