The specimen morgue of the metricians

Claim (verbatim)

Shams-i Qays al-Razi's al-Mu'jam fi ma'ayir ash'ar al-Ajam (المعجم فی معاییر اشعار العجم, early thirteenth century), edited by Qazvini, is the great medieval treatise on Persian prosody, rhyme and poetic criticism. Like every technical handbook it argues by example: to illustrate a meter, a rhyme-fault, a figure of speech, Shams-i Qays quotes a line or a couplet and names its author. Those shahids (evidentiary verses) are a peculiar preservation channel - poetry kept alive not for its beauty but for its usefulness as a specimen. The mechanism of loss is that a poet whose reputation could not carry a whole divan through the centuries could still have one serviceable line pinned in a prosodist's cabinet, exactly as a species known from a single museum specimen. A metrical treatise thus becomes an inadvertent morgue register of poets whose books are gone. If the treatise is doing its ordinary technical work rather than anthologizing the famous, a large share of the poets it cites as specimens should have no independent divan surviving to the print era, and therefore no corpus on Ganjoor. Prediction restated: among the named poets Shams-i Qays quotes as metrical or rhetorical examples, a large fraction exist today only as such specimens - no Ganjoor divan behind the cited line.

Prediction clause (verbatim)

Prediction: of the distinct named poets quoted as evidentiary verses in Shams-i Qays's al-Mu'jam (Qazvini edition), at least 40 percent will have no independent author corpus on Ganjoor - the cited specimen being effectively all that survives of them (primary clause: the specimen-only share at least 40 percent among al-Mu'jam's attributed shahids; the verdict follows it). Anonymous and formulaic examples are excluded; a poet counts as surviving only on an attributed Ganjoor poem-text; and the test voids for coverage if fewer than 60 distinct named poets can be extracted from the treatise's examples.

Kill-dataset (verbatim)

Kill: the printed al-Mu'jam of Shams-i Qays (Qazvini edition), listing every named poet cited as a prosodic or rhetorical example, each scored for presence or absence of an author corpus on Ganjoor (ganjoor.net). Distinct from the Yaqut Mu'jam al-buldan geographical instrument used elsewhere.

Novelty / leakage triage

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

Qazvini's edition of al-Mu'jam is the instrument, and the general phenomenon - technical treatises serving as sole witnesses for poets whose books are gone - is exactly why Lazard's corpus of early fragments draws on the shahid literature. But no published register scores al-Mu'jam's cited poets for survival against any corpus, and the >=40-percent specimen-only share against Ganjoor is un-run arithmetic. Nearest art is the edition itself and Browne's account of the treatise.

• Shams-i Qays, al-Mu'jam fi ma'ayir ash'ar al-'ajam, ed. Mirza Muhammad Qazvini (E.J.W. Gibb Memorial Series, Leiden-London, 1909)
• E.G. Browne, A Literary History of Persia, vol. II (London, 1906), on Shams-i Qays and the Mu'jam
• G. Lazard, Les premiers poetes persans (IXe-Xe siecles) (Tehran-Paris, 1964), drawing early-poet fragments from the technical literature

Predictions

No prediction registered yet.