The catalogue names more poets than the corpus keeps

Claim (verbatim)

Skaldatal - the catalogue of court poets that survives in the Uppsala manuscript of Snorri's Edda (DG 11) and in the paper apographs of the lost Kringla leaf - is a register of the profession itself: it lists the skalds by the rulers they served, from the legendary Danish and Swedish kings down to the Norwegian and Icelandic magnates of the thirteenth century, naming on the order of 140 poets. The Skaldic Project (skaldic.org) edits the surviving verse skald by skald. Set the two side by side and the catalogue becomes a survival experiment with its own control built in: it names the men remembered as court poets, and the database records whose verse a saga, a king's life, or Snorri's poetics actually troubled to quote. The loss mechanism is that skaldic verse survived almost only by being cited - a poet whose stanzas no prose author needed as evidence or specimen kept the catalogue's memory of his existence but not a line of his work. So a large fraction of Skaldatal's named poets should be zero-stanza ghosts: a name attached to a king, with nothing under it. Restated: at least two in five of the skalds Skaldatal names have no surviving whole stanza attributed to them in the Skaldic Project. The register counted the poets; the corpus kept only the citations.

Prediction clause (verbatim)

Prediction: matching the roughly 140 poets named in Skaldatal against the per-skald corpora of the Skaldic Project (skaldic.org), at least 40 percent of the catalogued skalds will have zero whole stanzas edited under their name (primary clause: the >=0.40 zero-stanza fraction; the verdict follows it). A skald counts as surviving only on at least one edited whole stanza attributed to him, not a half-line, a name in a prose sentence, or a bare lausavisa attribution without text; legendary pre-historical entries are retained in the denominator and flagged separately. The test voids for coverage if fewer than 110 Skaldatal entries resolve to a checkable poet-name.

Kill-dataset (verbatim)

Kill: the Skaldic Project database (skaldic.org) poet index, cross-matched name by name against the Skaldatal roster as edited from the Uppsala Edda (DG 11) and the Kringla apographs (in Finnur Jonsson's Edda Snorra Sturlusonar and the SkP introductory matter) - count the catalogued skalds carrying zero edited whole stanzas.

Novelty / leakage triage

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

Skaldatal's roster and the surviving corpus are both fully published - Jon Sigurdsson's edition and commentary in the Arnamagnaean Edda Snorra Sturlusonar annotates the catalogued skalds, Finnur Jonsson's Skjaldedigtning and now the Skaldic Project census the preserved verse, and Gudrun Nordal discusses the catalogue's poets who left no surviving stanzas. But no one has run the name-by-name zero-stanza fraction against the modern database. That many Skaldatal names carry no verse is a commonplace; whether the fraction clears 0.40 on skaldic.org's poet index, with the legendary entries retained in the denominator, is genuinely open and plausibly near the threshold rather than safely past it. Materials in print, arithmetic un-run.

• Edda Snorra Sturlusonar: Edda Snorronis Sturlaei, vol. III (Copenhagen: Legatum Arnamagnaeanum, 1880-87), Skaldatal with Jon Sigurdsson's commentary
• Finnur Jonsson, Den norsk-islandske skjaldedigtning, A I-II and B I-II (Copenhagen, 1912-1915)
• Gudrun Nordal, Tools of Literacy: The Role of Skaldic Verse in Icelandic Textual Culture of the Twelfth and Thirteenth Centuries (Toronto, 2001)
• D. Whaley (ed.), Poetry from the Kings' Sagas 1, Skaldic Poetry of the Scandinavian Middle Ages I (Turnhout, 2012), introduction

Predictions

No prediction registered yet.