A book that exists only in the footnotes

Claim (verbatim)

Saemundr Sigfusson inn frodi (1056-1133), the priest of Oddi, wrote a work on the kings of Norway - most probably in Latin - that no manuscript preserves. Its existence is not a modern inference but an explicit memory of the medieval writers who used it: Oddr Snorrason names Saemundr in his saga of Olafr Tryggvason, the vernacular synoptic histories draw on his reckoning, and Noregs konungatal, a late-twelfth-century poem, versifies the regnal chronology it credits to him, running its count of kings and reign-lengths back through Saemundr's authority. The loss mechanism is characteristic of the earliest Icelandic learning: a Latin scholarly history had no vernacular saga-audience to keep recopying it once its data had been absorbed into the sagas and poems that superseded it - the child texts cannibalised the parent, and then the parent starved. But the citations leave a provable shape: a chronology transmitted concordantly by several independent later works, with no surviving common source, is a phantom node whose existence, and even part of whose content, can be triangulated. Restated: Saemundr's lost history is attested and partly reconstructable from at least three mutually independent extant witnesses that name him or demonstrably carry his regnal reckoning, though the work itself survives in not one manuscript. A book can be counted precisely by the footnotes of the books that replaced it.

Prediction clause (verbatim)

Prediction: at least three mutually independent extant medieval works will be shown either to name Saemundr inn frodi as their authority for the succession of the Norwegian kings or to transmit a regnal chronology demonstrably derived from his reckoning, while no manuscript of the work itself survives (primary clause: at least three independent citing-or-deriving witnesses of a non-extant source; the verdict follows it). Independence requires that no two counted witnesses simply copy one another; Noregs konungatal, Oddr Snorrason's Olafs saga, and the synoptic histories are the candidate set. The test voids if fewer than three witnesses can be established as mutually independent.

Kill-dataset (verbatim)

Kill: the Islenzk fornrit apparatus and the standard editions of Noregs konungatal, Oddr Snorrason's Olafs saga Tryggvasonar, and the synoptic histories (Agrip and the Historia-tradition) - verify the independent citations of, or derivations from, Saemundr's lost konungatal.

Novelty / leakage triage

already answered in the literature

This is the established result of the source criticism of the earliest kings' histories, not an open count: Noregs konungatal states outright that its regnal reckoning down to Magnus godi follows Saemundr, Oddr Snorrason's Olafs saga Tryggvasonar cites him by name, and Ellehoj's monograph - the standard treatment - demonstrates the further derivations in the synoptic tradition while confirming that no manuscript of the work survives. The at-least-three-independent-witnesses structure the prediction asks for is exactly what Ellehoj (with Bjarni Adalbjarnarson before him and Andersson after) prints. The SkP II edition of Noregs konungatal restates the Saemundr dependence in its introduction, so the primary clause is already the published consensus.

• S. Ellehoj, Studier over den aeldste norrone historieskrivning (Bibliotheca Arnamagnaeana 26, Copenhagen, 1965)
• Bjarni Adalbjarnarson, Om de norske kongers sagaer (Oslo, 1937)
• T.M. Andersson, The Growth of the Medieval Icelandic Sagas (1180-1280) (Ithaca, 2006), on the pioneers Saemundr and Ari
• Noregs konungatal, in K.E. Gade (ed.), Poetry from the Kings' Sagas 2, SkP II (Turnhout, 2009)

Predictions

No prediction registered yet.