The shield-poems are only their captions now

Claim (verbatim)

The oldest and most ambitious skaldic long-poems are the ekphrastic shield- and mythological drapur - Bragi Boddason's Ragnarsdrapa, Thjodolfr of Hvinir's Haustlong, Ulfr Uggason's Husdrapa, Eilifr Godrunarson's Thorsdrapa - poems that once narrated whole mythological scenes at length. Almost none of them survives as a poem. They survive as the stanzas and half-stanzas that Snorri and the authors of the grammatical treatises pulled out to illustrate a kenning or a myth, embedded in Skaldskaparmal and the grammatical treatises, and nowhere else. The mechanism is the same citation-filter that governs the whole corpus, but concentrated at its most prestigious and most vulnerable point: a dense mythological drapa had no prose saga to ride in - no king's life needed it as evidence - so it survived only to the extent that a poetics anthology quarried it for parts, and only in the order and quantity the anthologist's argument required. What we call Husdrapa or Haustlong is therefore a reconstruction from quotations, not a transmitted text, and each is a fraction of the poem its own internal references imply. Restated: at least three of the canonical early ekphrastic drapur survive only through poetics-treatise quotation, with no independent or saga transmission, each preserved at well under its implied original length. The grandest early poems came down to us as the captions under Snorri's examples.

Prediction clause (verbatim)

Prediction: of the canonical early ekphrastic and mythological drapur (Ragnarsdrapa, Haustlong, Husdrapa, Thorsdrapa and their peers), at least three will be shown to survive solely through quotation in the poetics and grammatical treatises (Snorra Edda and the grammatical treatises) with no independent or prose-saga transmission channel (primary clause: at least three treatise-only early drapur; the verdict follows it). Each such poem's surviving extent, counted in whole stanzas edited in the Skaldic Project, is expected to fall well short of the length its own narrative scope implies - a secondary observation, flagged as qualitative rather than load-bearing. The test voids if the prose carrier field for these poems cannot be read in skaldic.org.

Kill-dataset (verbatim)

Kill: the Skaldic Project database (skaldic.org), reading each named early drapa's recorded prose carrier(s) - confirm the poems whose every surviving stanza is transmitted only via Snorra Edda and the grammatical treatises, with no saga or otherwise independent witness.

Novelty / leakage triage

already answered in the literature

The transmission channel is stated poem by poem in the Skaldic Project's own volume: the SkP III introductions to Ragnarsdrapa, Haustlong, Husdrapa and Thorsdrapa record that every surviving stanza of each is transmitted in Snorra Edda and the grammatical-treatise tradition, with no saga or other independent carrier - Laxdaela saga narrates Husdrapa's composition but quotes not a line of it. Turville-Petre's handbook edition and Faulkes's Skaldskaparmal state the same for all four, so the >=3 threshold is guaranteed four times over before any query is run. The far-short-of-implied-length observation is likewise standard in those introductions; nothing load-bearing is left to count.

• K.E. Gade and E. Marold (eds.), Poetry from Treatises on Poetics, Skaldic Poetry of the Scandinavian Middle Ages III (Turnhout, 2017), poem introductions to Ragnarsdrapa, Haustlong, Husdrapa and Thorsdrapa
• E.O.G. Turville-Petre, Scaldic Poetry (Oxford, 1976)
• A. Faulkes (ed.), Snorri Sturluson, Edda: Skaldskaparmal (London, 1998)

Predictions

No prediction registered yet.