Ars Inquirendi

AI-generated conjecture · below the evidence/publication boundary

← All conjectures · Works' qualities & composition

The poems jump the frame

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)

Al-Hariri's Maqamat is a prosimetrum whose two phases age differently in other people's books. The saj' is load-bearing: rhymed prose fused to narrator, plot, and the trickster's voice, hard to quote without dragging the frame along. The embedded verses are the opposite: metrical, self-bounded, address-free — prefabricated for extraction into anthologies, commentaries, and epistles. A mixed-form work sheds its verse into the quotation stream while its prose stays home, and the shedding should accelerate as living memory of the frame recedes. Prediction: in KITAB reuse alignments from the Maqamat into later OpenITI works, verse passages will be reused at a per-word rate at least three times that of the prose (primary clause: the threefold verse-to-prose per-word reuse ratio; the verdict follows it), and the ratio will rise, not fall, with the chronological distance of the reusing work from al-Hariri. Exact computation: segment the OpenITI Maqamat text into verse and prose spans, intersect the spans with KITAB passim alignment offsets, and compute per-word reuse rates by span type and by reuser century. Kill: the KITAB pairwise text-reuse data and the OpenITI text of al-Hariri's Maqamat, both openly downloadable on GitHub.

Prediction clause (verbatim)

Prediction: in KITAB reuse alignments from the Maqamat into later OpenITI works, verse passages will be reused at a per-word rate at least three times that of the prose (primary clause: the threefold verse-to-prose per-word reuse ratio; the verdict follows it), and the ratio will rise, not fall, with the chronological distance of the reusing work from al-Hariri. Exact computation: segment the OpenITI Maqamat text into verse and prose spans, intersect the spans with KITAB passim alignment offsets, and compute per-word reuse rates by span type and by reuser century.

Kill-dataset (verbatim)

Kill: the KITAB pairwise text-reuse data and the OpenITI text of al-Hariri's Maqamat, both openly downloadable on GitHub.

Provenance

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

Fresh blind generation instance of claude-fable-5, 2026-07-16, wave M02 (the works) of the Minds & Works campaign, produced from model knowledge alone under the two-file blindness protocol.

Novelty / leakage triage

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

The extractability of al-Hariri's embedded verse vs his load-bearing saj' is a plausible commonplace of maqama reception, but the specific KITAB prediction (verse reused >=3x the prose per-word, ratio rising with chronological distance) was not located.

  • KITAB project text-reuse datasets; OpenITI text of al-Hariri's Maqamat

Predictions

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

Resolution: Killed

Caveats: ROBUSTLY KILLED. Al-Hariri's embedded VERSE is reused at essentially the SAME per-word rate as his saj' PROSE (R = 1.04), not the >=3x the conjecture staked. The premise -- that self-bounded verse is prefabricated for anthology extraction while the frame-fused rhymed prose stays home -- does not hold for the Maqamat, whose saj' is itself its celebrated, endlessly-quoted feature, so the prose travels about as readily as the verse. A verse/prose line-classification bug (continuation lines wrongly inheriting a poem's status across a wrap into prose) was caught and fixed mid-build and validated. Honest caveat, non-verdict-changing given the distance from 3x: some counted 'reuse' comes from very late (even modern) targets quoting this famous text as a rhetorical exemplar rather than medieval recombination; the registered century-trend secondary is noisy and non-monotonic (no clean rise with chronological distance). Same KITAB pipeline independently validated via w02-010.

Registered against KITAB reuse data over OpenITI. Claim under test (primary clause): in reuse flowing OUT of al-Hariri's Maqamat, the embedded VERSE is reused at a per-word rate AT LEAST THREE TIMES that of the saj'/PROSE frame; secondarily the verse-to-prose ratio rises, not falls, with the reusing work's chronological distance from al-Hariri.

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: KITAB alignments whose SOURCE side falls in al-Hariri's Maqamat, target = any later OpenITI work. SEGMENTATION: VERSE spans = spans the OpenITI Maqamat text marks as poetry (mARkdown verse tags, e.g. the '# ' poetry / hemistich '%~%' markup); PROSE spans = the remaining running text. RATE(type) = (reused words landing in spans of that type) / (total words of that type). PRIMARY R = RATE(verse) / RATE(prose). CLAUSE PRECEDENCE: (1) INCONCLUSIVE_BY_DESIGN if the Maqamat is absent, OR the text carries no poetry markup to segment verse from prose, OR total reuse instances from the Maqamat < 20. (2) SUPPORTED if R >= 3. (3) KILLED if < 3. Report RATE(verse), RATE(prose), R, verse/prose word totals; the reuser-century trend of R is a registered SECONDARY (non-binding) diagnostic.

Known priors disclosure: Held: the conjecture's reasoning (rhymed-prose saj' is fused to frame/voice and resists extraction; embedded verse is self-bounded and prefabricated for anthologizing) and general knowledge the Maqamat is in OpenITI with poetry markup and KITAB coverage. NOT computed: the per-span reuse rates or their ratio. Threefold verse:prose 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). 1,065 outgoing alignments (100% to later works). Verse/prose classified per-line by content (a line-wrap-inheritance bug was caught + fixed mid-build).

Dataset: KITAB reuse instances flowing OUT of al-Hariri's Maqamat (into later works), with the Maqamat text segmented into VERSE spans (# poetry lines carrying the caesura pattern) vs PROSE spans. R = per-word reuse rate(verse) / rate(prose).

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.