The oldest Korean history is built on older ones it replaced

Claim (verbatim)

The Samguk sagi (三國史記, Samguk sagi, 1145, Kim Busik) is the oldest surviving Korean history, and it is a survival census in its own right, because it names its sources - and almost all of them are gone. It cites earlier native records now wholly lost: the Silla Guksa (國史, compiled 545 by Geochilbu 居柒夫), the Goguryeo Sinjip (新集, 600, itself a condensation of the older Yugi 留記), the Baekje Seogi (書記, by Go Heung 高興), the Hwarang segi (花郎世記, Kim Daemun 金大問), and assorted gogi (古記, old records). Behind Kim Busik's own work stands the Gu samguksa (舊三國史, 'Old History of the Three Kingdoms'), itself lost, but proven to have existed because Yi Kyubo (李奎報) quoted its Dongmyeongwang bon-gi (東明王本紀) in his Dongmyeongwang-pyeon (東明王篇) of 1193. The mechanism is founder-supersession on a peninsular scale: a new authorized history cannibalizes and then displaces the chronicles it draws on, and a small state's thin manuscript ecology cannot keep the superseded copies alive. Prediction restated: among the distinct earlier sources the Samguk sagi names, the non-extant outnumber the independently extant by at least ten to one. The founding history of Korea is, in bibliographic terms, a graveyard of the histories it was made from.

Prediction clause (verbatim)

Prediction: enumerating the distinct earlier text-sources named within the Samguk sagi (三國史記), the number that are non-extant (surviving only as names, or as quotations inside the Samguk sagi and a few later works) will exceed the number independently extant by at least ten to one (primary clause: the >=10:1 non-extant-to-extant ratio among named earlier sources; the verdict follows it); a source counts as extant only on an independent surviving text, and the lost Gu samguksa (舊三國史) is scored as attested-but-lost on the strength of Yi Kyubo's quotation.

Kill-dataset (verbatim)

Kill: the Samguk sagi (三國史記) full text in the National Institute of Korean History database (db.history.go.kr), enumerating the cited earlier source-titles and scoring each for independent survival, with Yi Kyubo's Dongmyeongwang-pyeon (東明王篇) confirming the attested-but-lost Gu samguksa.

Novelty / leakage triage

already answered in the literature

The discipline flags this correctly: Samguk sagi source criticism has catalogued exactly this. Yi Gang-rae's 三國史記 典據論 enumerates the earlier native records Kim Busik names — the Silla Guksa, Goguryeo Sinjip/Yugi, Baekje Seogi, Hwarang segi, the gogi, and behind them the Gu samguksa (attested through Yi Kyubo's Dongmyeongwang-pyeon) — and establishes that essentially all are lost while independently surviving native sources are a bare handful. The >=10:1 non-extant-to-extant ratio among named sources is thus the standing result of the source-critical literature, not an open test. Leaked.

• 이강래 (Yi Gang-rae), 삼국사기 전거론 (三國史記 典據論) (Seoul: Minjoksa, 1996)
• 李奎報 (Yi Kyubo), 東明王篇 (1193), quoting the 東明王本紀 of the lost 舊三國史

Predictions

No prediction registered yet.