Ars Inquirendi

← All conjectures · Islamicate book cultures

The wandering quatrains

Status: Anticipated in print · untested

The verdict’s fine print — quoted from the resolution record: “CORPUS VINTAGE/SCOPE (the whole ruling): the only published Route A export is Ganjoor's pre-rewrite 2014-10-03 snapshot with 67 of the live site's 233 poets, so C covers ~29% of the poets the registrant was looking at when the wager was placed; N=17 is a lower…” Read the full caveats ↓

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.
“Anticipated in print” means the direction of this claim already appears in scholarship, while this exact test has never been run — it records prior art, not proof; nothing here counts as supported or falsified until a registered resolution says so.

Claim (verbatim)

The ruba'i, the Persian quatrain, is the most portable and least anchored unit of the tradition: four lines, easily memorized, easily detached from any author, easily re-attributed. Around Umar Khayyam (d. c. 517/1123) this produced the tradition's most notorious attribution cloud. Khayyam was famous as a mathematician and philosopher; over the centuries the number of quatrains ascribed to him swelled from a few dozen in the early sources to well over a thousand in the late collections, and the surplus was not composed but recruited - quatrains that also circulate under Attar, Iraqi, Baba Afdal, Abu Sa'id, and other names drifted onto the most magnetic available signature. The mechanism is orphan-adoption in a portable genre: a free-floating quatrain acquires a famous father, and the same four lines end up filed under two or three poets at once. Ganjoor, built faithfully from separate printed divans, does not silently merge these - it hosts the Khayyam collection and the other poets' quatrains side by side, so the overlaps are directly countable as near-identical quatrains sitting under different names. Prediction restated: a substantial, countable set of Ganjoor's Khayyam quatrains appear, as substantially the same quatrain, under at least one other poet in the same corpus - the wandering quatrains caught in the act of being in two places at once.

Prediction clause (verbatim)

Prediction: at least 30 distinct quatrains on Ganjoor's Khayyam (خیام) page will also appear, as substantially the same quatrain (matching on rhyme-words and at least three of four misra'), under at least one other Ganjoor poet's corpus (primary clause: the count of cross-attributed Khayyam quatrains at least 30; the verdict follows it). Matching is on near-identical text, not theme; trivial variants count as the same quatrain; and the test reports the partner poets so the direction of drift can be inspected. The threshold is an absolute count set well clear of chance, robust to the exact size of Ganjoor's Khayyam collection.

Kill-dataset (verbatim)

Kill: the Ganjoor (ganjoor.net) corpus - compare the Khayyam ruba'iyat against the ruba'iyat of all other Ganjoor poets (its own similar-poems / اشعار مشابه index or a direct near-duplicate match over the quatrain texts), counting Khayyam quatrains that recur under another name.

In the kill-dataset registry:

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 scholarship. Kills and prior scholarship are credited here, by name, as they come in.

Provenance

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

Fresh blind generation by claude-fable-5, 2026-07-17, Persianate wave instrument-anchored on the open Ganjoor corpus (ganjoor.net) with the Cambridge Shahnama Project as the second instrument where manuscript transmission is the question. Every Kill names one corpus and one countable/positional/citation-geometric operation, thresholds sit far from 1, disambiguation and name-matching rules are pinned inside each prediction, and coverage guards separate what Ganjoor holds (print-era critical editions) from what existed. DISJOINTNESS: the Nizami/Khamsa transmission ground and de Blois vol. V's Nizami dated-manuscript census are owned by minds-w02 (item 'The Khamsa is bound, not born') and are NOT re-posed here; the Rumi Masnavi return-formula operation on Ganjoor is owned by minds-w02 and avoided; the Ferdowsi age-statement / reign-boundary operation on the Ganjoor Shahnameh is owned by minds-w01 ('The Shahnameh keeps the poet's birthdays') and the Shahnama interpolation-topography / illustrated-extent / folio-grid / dispersal / defacement operations are owned by fable-w01 (#14), fable-w03 (#24) and fable-w05 (#3,#4,#11) on the Cambridge Shahnama Project - so the Daqiqi item here reuses the Ganjoor Shahnameh and Cambridge Shahnama instruments ONLY under a distinct survival-by-embedding operation, flagged in its Kill; prose Kalila wa Dimna transmission is owned by fable-w01 (#22) and fable-w07 (#11,#22), so the verse-Kalila point is folded into the Rudaki item and no standalone Kalila/Sindbad-nama item is posed. Safa's Tarikh-i adabiyat and de Blois' Bio-Bibliographical Survey are reused as census controls under attested-poet/attested-title operations distinct from the minds-w02 Nizami dated-ms count. NOTE cleared: the 'Mu'jam' in islamicate-openiti (#4) is Yaqut's geographical Mu'jam al-buldan, a different work from Shams-i Qays's prosody al-Mu'jam used here - no collision. DROPPED CANDIDATES: (a) Nizami's own lists / khamsa-imitation chains of lost romances - dropped as Nizami is owned by minds-w02; recast as the non-Nizami verse-romance census (item 12). (b) A standalone Sindbad-nama / verse-Kalila item - dropped, prose Kalila owned by fable-w01/w07; the verse-Kalila loss folded into item 1. (c) A raw-bayt genre-size comparison (three Ghaznavid panegyrists vs one lyricist) - dropped as FALSE on the numbers (Farrukhi's surviving divan alone is large), replaced by the within-Ganjoor pre-Mongol survival-concentration op (item 6). (d) A smooth recency-gradient census - replaced by the mechanistic Mongol-caesura step (item 4). CONFIDENCE FLAGS carried inside the items: Rudaki's 100,000-couplet attestation and 'Unsuri's ~30,000-bayt divan are classical topoi (kept out of the load-bearing clauses, which rest on the hard surviving counts); the exact women-poet roster and the Khayyam cross-attribution magnitude are MODERATE and primary clauses are phrased to survive the uncertainty.

Novelty / leakage triage

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

The analog answer is emphatically in print: Zhukovsky's wandering-quatrains study identified 82 Khayyam-attributed quatrains also transmitted under other names, Christensen's two monographs enlarged and systematized the cloud, and Aminrazavi surveys it - so a cross-attribution count well above 30 exists in the literature for the manuscript tradition. What is not in print is the Ganjoor-internal count the primary clause pins: Ganjoor's Khayyam page descends from curated printed selections (the Forughi-Ghani 178) assembled precisely to exclude suspect attributions, so whether >=30 of its quatrains recur near-verbatim under other Ganjoor poets is a genuine, un-run question about the digital corpus rather than a restatement of Zhukovsky. Adjacent, with the caveat that the phenomenon itself is as thoroughly published as a phenomenon can be.

Sources cited by the triage
  • V.A. Zhukovsky, ''Umar Khayyam and the Wandering Quatrains' (1897), Eng. trans. E.D. Ross, Journal of the Royal Asiatic Society 30 (1898), identifying 82 cross-attributed quatrains
  • A. Christensen, Recherches sur les Ruba'iyat de 'Omar Hayyam (Heidelberg, 1905), and Critical Studies in the Rubaiyat of 'Umar-i-Khayyam (Copenhagen, 1927)
  • M. Aminrazavi, The Wine of Wisdom: The Life, Poetry and Philosophy of Omar Khayyam (Oxford, 2005)
  • Ruba'iyyat-i Khayyam, ed. M.'A. Forughi and Q. Ghani (Tehran, 1321sh/1942), the curated 178-quatrain selection behind the standard digital texts

Predictions

Inconclusive registered 2026-07-19 variant: v1 calibration prediction (parent triage: leaked/adjacent)

Resolution: Inconclusive

Caveats: CORPUS VINTAGE/SCOPE (the whole ruling): the only published Route A export is Ganjoor's pre-rewrite 2014-10-03 snapshot with 67 of the live site's 233 poets, so C covers ~29% of the poets the registrant was looking at when the wager was placed; N=17 is a lower bound on the registered quantity, not a measurement of it. WHY ENUMERABLE ROUTE A DOES NOT FORCE CLAUSE (3): clause (3) presupposes N measures the registered quantity; the registration's own verification evidence (live site, live API) fixes the referent corpus as current Ganjoor, and Route A's silent decade of vintage-drift was discoverable only at compute time -- killing on the subset would let an instrument's undisclosed staleness convert a lower bound into a refutation, the exact false-kill mode lesson 7 guards against (same void family as romanlaw-011 and greekastro-009/013). THE CONCENTRATION ARGUMENT, WEIGHED AND REJECTED AS A KILL LICENSE: yes, all 17 matched classes partner with major classical rubaiyat poets present in the snapshot (Attar 8, Abu Sa'id 3, Rumi 2, Sanai/Baba Afzal/Khaqani/Shah Nematollah 1 each), but that observation is circular about absent poets, and the shepherd's roster check shows the 166 absent poets are NOT overwhelmingly minor/modern: live Ganjoor carries rubaiyat sections for snapshot-absent classical poets squarely in the wandering-quatrain stratum -- e.g. Kamal al-Din Esmail, Awhad al-Din Kermani (a corpus on the order of 1,700 quatrains), Zahir Faryabi, Kamal Khojandi, Ibn Yamin, Mojir Beylaqani -- so the +13 NEW distinct Khayyam classes needed to reach 30 (out of 238 currently unmatched classes) are plausibly reachable and a kill here is not safe. HONEST WEIGHT ON THE OTHER SIDE (not suppressed): Route A was literally enumerable and sha256-clean; Khayyam is over-represented in it (320 vs the assumed 178), which inflated, not deflated, matching chances; and on the delivered instrument the wager fell 43% short (17 vs 30) under a rule the registrant self-declared conservative -- direction confirmed (17 genuine wandering classes, spot-verified variants), magnitude overshot, the campaign's recurring calibration signature. This void does NOT vindicate the conjecture: V2-ELIGIBLE with a re-pinned instrument (current-corpus enumeration -- e.g. the live gdb/API with a feasible enumeration budget, or a fresh full-corpus export if Ganjoor publishes one); if v2 on the full 233-poet corpus still lands under 30, that is a clean kill. Reported-non-verdict per registration: partner-poet table as above; |Q|=320; within-Q duplication 0.403 (Ganjoor's Khayyam section itself heavily self-duplicating across its Robaiyat and Hedayat sub-trees).

Registered against Ganjoor (poet خیام نیشابوری confirmed present at registration, id=3, fullUrl /khayyam; instrument permission as verified for persianate-ganjoor-005: robots on both hosts fully permissive, API live). PREDICTION VERBATIM: "Prediction: at least 30 distinct quatrains on Ganjoor's Khayyam (خیام) page will also appear, as substantially the same quatrain (matching on rhyme-words and at least three of four misra'), under at least one other Ganjoor poet's corpus (primary clause: the count of cross-attributed Khayyam quatrains at least 30; the verdict follows it). Matching is on near-identical text, not theme; trivial variants count as the same quatrain; and the test reports the partner poets so the direction of drift can be inspected. The threshold is an absolute count set well clear of chance, robust to the exact size of Ganjoor's Khayyam collection." FULL OPERATIONALIZATION PINNED AT REGISTRATION (lesson 3 - every matching choice frozen here): CORPORA: Q = all poems of poet id 3 whose verse structure is exactly 4 misra; C = all poems with exactly 4 misra of every other poet (id != 3). ROUTE: preferred route A = Ganjoor's official bulk database exports (the project's published per-poet SQLite/gdb files), one download per poet; fallback route B = the public API, sequential with >=1s self-imposed spacing; the compute documents which route and every URL/endpoint. NORMALIZATION (pinned, applied to every misra): Unicode NFC; character maps ي>ی, ك>ک, ة>ه, ۀ>ه, أ>ا, إ>ا, ٱ>ا, ؤ>و, ئ>ی; strip combining marks U+064B-U+0655 and U+0670; strip tatweel U+0640; delete ZWNJ U+200C; strip Arabic and Latin punctuation; collapse whitespace. RHYME-WORD = the final whitespace token of the normalized misra. SIMILARITY = Python difflib.SequenceMatcher(None, a, b).ratio() on the two normalized misra strings (function pinned exactly). MATCH RULE (pinned): quatrains q and c MATCH iff, over the four positionally aligned misra (no permutation - conservative, disclosed), at least 3 of 4 positions have similarity >= 0.75 AND at least 3 of 4 positions have identical rhyme-words. This double condition is the registered realization of 'matching on rhyme-words and at least three of four misra'' with 'trivial variants count as the same quatrain'. DISTINCTNESS (pinned): Khayyam-internal equivalence classes are formed by union-find over the same MATCH rule; the counted unit is the equivalence class. PRIMARY QUANTITY N = number of distinct Khayyam quatrain classes with at least one MATCH to a quatrain of a different poet.

Resolution criteria — the registered fine print

Resolution criteria: COMPUTE RECIPE (blind agent): (1) obtain the full Ganjoor corpus by route A or B (documented); (2) extract Q and C by the exactly-4-misra structure rule; (3) normalize, compute MATCHes under the pinned rule (an efficient blocking scheme - e.g. prefiltering on shared rhyme-words - is permitted ONLY if provably equivalent to exhaustive comparison under the pinned rule, and must be stated); (4) form Khayyam-internal classes, count N, and tabulate partner poets per matched class. CLAUSE PRECEDENCE: (1) INCONCLUSIVE_BY_ACCESS only on instrument failure (the Khayyam section or the other-poet corpus cannot be enumerated by either route) - the prediction pins no measured-value void, so none is added; (2) SUPPORTED if N >= 30; (3) REFUTED if N < 30. REPORTED, NON-VERDICT: the partner-poet table (the prediction's own reporting requirement, 'so the direction of drift can be inspected'), the size of Q, and the within-Q duplication rate. Compute firewalled from the threshold and from this disclosure.

Known-priors disclosure — what the registrant already knew

Known priors disclosure: Triage 2026-07-17 graded ADJACENT with the strongest possible phenomenon-level prior disclosed: Zhukovsky 1897 identified 82 'wandering quatrains' among Khayyam-attributed rubaiyat in the manuscript tradition, enlarged by Christensen 1905/1927 and surveyed by Aminrazavi 2005 - so a cross-attribution count far above 30 exists in print FOR THE MANUSCRIPT TRADITION. What is un-run (triage's grounds for adjacency) is the Ganjoor-internal count: Ganjoor's Khayyam section descends from the curated Forughi-Ghani 178, assembled precisely to EXCLUDE suspect attributions, so whether >= 30 of ITS quatrains recur under other GANJOOR corpora (Attar's Mokhtarnameh, the Abu-Sa'id-attributed quatrains, Baba Afzal, the Divan-e Shams rubaiyat, etc., as hosted) is a genuine question about the digital corpus. Registrant expectation genuinely open: the curation cuts one way, the sheer size of Ganjoor's quatrain holdings the other. CALIBRATION NOTE: absolute-count pin, self-described as 'set well clear of chance' - not the overpinned multiplicative class. The pinned 0.75/3-of-4 positional rule is conservative (no hemistich permutation), which biases against SUPPORT if anything - disclosed ex ante. PROBE DISCLOSURE: only the poet roster was fetched at registration; no quatrain text was seen.

Method and dataset — how it was measured

Firewalled blind compute (claude-sonnet-5, blind to the N>=30 threshold and to direction) implemented the pinned rule exactly: NFC + pinned char-map normalization, final-token rhyme words, rhyme-blocked candidate generation proven lossless vs all-pairs, MATCH iff >=3/4 positionally-aligned misra with SequenceMatcher ratio >= 0.75 AND >=3/4 identical rhyme words (no permutation), union-find distinctness over Q, N = distinct Khayyam classes with >=1 external match. The Opus orchestrator independently re-implemented the entire pinned rule from scratch and reproduced N=17 EXACTLY; spot-checks confirmed the matches are genuine wandering-quatrain textual variants (e.g. Khayyam robai 2 vs Attar Mokhtarnameh 44). Verdict authored by claude-fable-5 per the registered clause precedence: clause (1) INCONCLUSIVE_BY_ACCESS fires because the REGISTERED other-poet corpus -- every other poet of the Ganjoor the registrant demonstrably pinned (233 poets, live-verified at registration) -- cannot be enumerated by either route: Route A's only existing export delivers a 2014 67-poet predecessor corpus, and Route B at the pinned >=1s spacing cannot exhaustively enumerate 233 poets' corpora in bounded time. N=17 on the deliverable subset is a LOWER BOUND on the registered quantity and clauses (2)/(3) are therefore not reached. Per the registration's own void rule, no measured-value void is added. VERDICT: INCONCLUSIVE.

Dataset: Route A as pinned: Ganjoor's official published bulk SQLite export -- the ONLY such export in existence, SourceForge s3db/ganjoor-s3db-930711.zip (Persian date 930711 = 2014-10-03 CE), sha256 cb782243a10824ca9247dc6529d32326df77f840a3be510c2cdb7fade5071866 verified independently by compute agent and shepherd. It is a PRE-REWRITE snapshot covering 67 poets; the live ganjoor.ir that the registrant verified at registration (poet id=3 at /khayyam, API live) carries 233 poets, so the delivered instrument is a ~29%-of-poets, decade-stale subset of the registered corpus. Q = poet id 3 (خیام, verified) exactly-4-misra poems: 320 (vs the assumed curated 178 -- Khayyam is OVER-represented, including Hedayat's Taraneh-ha sub-tree). C = all other poets' exactly-4-misra poems: 11,873. Shepherd re-verified poet count, |Q|, |C| directly against the DB and enumerated the full 67-poet roster for the corpus-scope ruling.

computed 2026-07-20

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.