whose.one source dossier · v0.4 · 2026-08-24

Liancourt Rocks — Dokdo · Takeshima
Source Dossier

The documentary evidence behind both official narratives, with the accessibility of each original noted. Every source carries the other side's rebuttal. This document endorses neither claim.

X Facebook
VOTE Which story do you find more convincing? Naming rule This is the English edition, so the neutral name leads (Dokdo leads in Korean, Takeshima in Japanese) · The order of the two narrative sections below is randomized on every visit
cited by Koreacited by Japanneutral fact

Current status — as of August 2026 Status as of 2026-08

Change log

Japan's defence white paper again described the islets as inherent Japanese territory with the dispute unresolved; South Korea's defence ministry summoned the Japanese defence attaché and demanded the wording be corrected.

The Diplomatic Bluebook 2026, reported to cabinet, repeated that South Korea occupies the islets with no basis in international law, and recorded Japan's protest over South Korean military exercises there in 2025.

Japan's education ministry cleared high-school textbooks describing the islets as inherent Japanese territory illegally occupied by South Korea; Seoul summoned the Japanese embassy's minister and protested.

The last permanent civilian resident died; the registered residential population is now zero.

The lighthouse was upgraded from unmanned to manned.

The order of the two narrative sections below is randomized on every visit — this dossier takes no side, not even in sequence.

Korea's narrative 6 sources

A chain of historical records (512→1454→1770) + Japan's own denials (1696·1877) + a modern act of sovereignty (1900) + postwar dispositions (1946).

Samguk Sagi — the submission of Usan-guk 512 · compiled 1145

于山國歸服 歲以土宜爲貢

Records that Silla's general Isabu subjugated Usan-guk, the maritime polity centered on Ulleungdo. Korea holds that Dokdo fell within Usan-guk's domain.

Meaning
The starting point of a governance lineage — predating Japan's earliest documented awareness of the island (1667) by over a millennium.
Japan rebuts
The record mentions only Usan-guk (Ulleungdo); no island identifiable as Dokdo appears. Subjugating Usan-guk does not in itself demonstrate rule over Dokdo.
original published documented Full text: National Institute of Korean History DB MOFA Korea — grounds

Geography section, Annals of King Sejong 1454

風日淸明 則可望見

"The two islands of Usan and Mureung lie due east of Uljin in the sea… on a clear day they are visible from one another." Korea reads Usan as Dokdo and Mureung as Ulleungdo — the only distinct island visible from Ulleungdo with the naked eye is Dokdo.

Meaning
A state-compiled geography placing the island under Korean administration (Uljin County).
Japan rebuts
'Usando' in old Korean texts is an alias of Ulleungdo or a non-existent island; on the map in the Sinjeung Dongguk Yeoji Seungnam, Usando is drawn west of Ulleungdo at a similar size — it cannot be Dokdo.
original published documented Veritable Records DBMOFA Japan rebuttal

The An Yong-bok affair and the Tottori reply 1693–1696

During the Ulleungdo dispute (Takeshima Ikken), the shogunate queried Tottori domain, which replied that "neither Takeshima (Ulleungdo) nor Matsushima (Dokdo) belongs to Tottori." The shogunate then banned Japanese crossings to Ulleungdo in January 1696.

Meaning
Japan's central government, after checking with its own domain, acknowledged in the 17th century that neither island was Japanese — directly contradicting the claim that sovereignty was "established by the mid-17th century."
Japan rebuts
① The 1696 ban covered only Ulleungdo; crossings to Takeshima (Dokdo) were never banned — evidence, Japan argues, that it regarded the island as its own. ② An Yong-bok's testimony was given under interrogation as a lawbreaker and is unreliable. — Korea's re-rebuttal: the Tottori reply is a Japanese official document independent of An's testimony.
original published documented Held by Tottori Prefectural Museum Korean History NetMOFA Japan rebuttal

The Dajōkan Directive + Isotakeshima sketch map 1877

竹島外一島之儀 本邦關係無之

Japan's highest state organ, the Dajōkan, ruling on a land-registry query from Shimane, directed: "Concerning Takeshima (Ulleungdo) and one other island, know that they have no relation to our country." The attached sketch map depicts that "other island" — Matsushima, i.e., Dokdo.

Meaning
An official confirmation by Japan's own modern central government that Dokdo was not Japanese — in tension (estoppel) with the 1905 incorporation's claim of "reconfirming inherent territory." The strongest single document in Korea's narrative.
Japan rebuts
Japan's Foreign Ministry site never mentions the directive (absent from the official narrative). Rebuttals come from Shimane's Takeshima研究会: "whether the 'other island' means Dokdo is unclear; the document dates from a period of confused island naming." Korean scholarship replies that the sketch map's explicit Matsushima defeats that reading.
磯竹島 松島 隱岐 →
Reproduction of the 1877 sketch map — the 'other island' (Matsushima/Dokdo) drawn beside Isotakeshima is the crux of Korea's reading
original published documented Held by the National Archives of Japan MOFA Korea — scanEncyclopedia of Korean Culture

Imperial Edict No. 41 of the Korean Empire 1900.10.25 · official gazette

鬱陵全島 竹島 石島

Elevated Ulleungdo to Uldo County and defined its jurisdiction as "all of Ulleungdo, Jukdo, and Seokdo (石島)." Korea identifies Seokdo as Dokdo: 'stone island' → dialectal dok island → 石島 by meaning, 獨島 by sound.

Meaning
A modern, gazette-published assertion of jurisdiction five years before Japan's 1905 incorporation — meaning the island was not terra nullius in 1905.
Japan rebuts
① If Seokdo were Dokdo, why not write 'Dokdo' or 'Usando'? (Some identify Seokdo as Gwaneumdo.) ② Even granting the identification, there is no evidence of actual control around 1900. — Korea's re-rebuttal: county head Sim Heung-taek's 1906 report ("Dokdo, under this county's jurisdiction") shows the edict in operation.
original published documented Gazette original at Kyujanggak MOFA KoreaMOFA Japan rebuttal

SCAPIN 677 · 1033 1946

Allied directive 677 (Jan 29) expressly excluded Ulleungdo, Liancourt Rocks, and Jeju from the area of Japanese governmental administration. Directive 1033 (Jun 22, the MacArthur Line) barred Japanese vessels from within 12 miles of the rocks.

…excluding (a) Utsuryo (Ullung) Island, Liancourt Rocks (Take Island) and Quelpart (Saishu or Cheju) Island… — SCAPIN 677, §3
Meaning
Immediately after the war the Allies treated the island as outside Japan's administrative reach — a status Korea says carried into its 1948 founding.
Japan rebuts
Both directives contain their own caveat — "nothing in this directive shall be construed as an indication of Allied policy relating to the ultimate determination" of territory (677 §6, 1033 §5). Final disposition, Japan argues, was made by the San Francisco treaty. Notably, MOFA Japan itself publishes the SCAPIN scans in full.
original published documented MOFA Japan (full PDFs)MOFA Korea FAQ

Japan's narrative 6 sources

A five-step structure: 17th-century inherent-territory thesis → 1905 reaffirmation of intent → postwar treaty settlement (SF Treaty + Rusk) → continuous administration → standing offer of ICJ adjudication.

Crossing permits and the Onshū records 1618/1625 · 1667

然則日本之乾地 以此州爲限矣

The shogunate licensed the Ōya and Murakawa families to sail to Ulleungdo; en route they used Matsushima (Dokdo) as a waypoint, anchorage, and fishing ground. The 1667 Onshū Shichō Gōki is cited as evidence of early Japanese cognizance.

Meaning
Core of the thesis that "sovereignty over Takeshima was established by the mid-17th century at the latest."
Korea rebuts
① Crossing permits were issued for voyages to places outside the realm — evidence, rather, that Japan saw the islands as foreign. ② In "Japan's northwest limit is this province (此州)," the province is Oki in context — placing both islands outside Japan (Korea's official reading). ③ Japan itself later denied ownership: the 1696 ban and the 1877 directive.
copies published documented Permit originals not extant Cabinet Secretariat commentaryMOFA Korea on the Onshū text

Cabinet decision · Shimane Notice No. 40 1905.1.28 / 2.22

Prompted by sea-lion entrepreneur Nakai Yōzaburō's petition (Sept 1904), the cabinet found "no trace of occupation by another power" (terra nullius) and approved incorporation; Shimane Prefecture issued the notice. Fishing licenses and land-registry entries followed.

Meaning
Both a "reaffirmation of intent over inherent territory" and a modern incorporation satisfying the international-law requirements of occupation — the starting point of effective control.
Korea rebuts
① Edict No. 41 preceded it by five years — the island was not terra nullius. ② Nakai himself initially believed the island Korean and planned to petition Seoul. ③ A prefectural notice, not a state gazette — no internationally effective notification. ④ Timing: mid Russo-Japanese War, months before the 1905 protectorate treaty stripped Korea of its diplomacy — protest was impossible. ⑤ 'Inherent territory' and 'terra nullius occupation' contradict each other. The notice's original burned in 1945; copies survive.
copies published documented Cabinet decision: JACAR MOFA Japan

San Francisco Peace Treaty, Art. 2(a) signed 1951.9.8

The final text names only three islands; Liancourt Rocks is silent. Draft history: US drafts of 1947–49 listed it among territories renounced to Korea → after political adviser Sebald's memoranda (Nov 1949) it moved to Japan's column → from the Aug 1950 draft it vanished from both lists → final silence.

Japan… renounces all right, title and claim to Korea, including the islands of Quelpart, Port Hamilton and Dagelet. — Art. 2(a)
Meaning
Japan's official reading: Korea asked for the island to be added, the US refused (the Rusk letter), therefore the treaty affirms it as Japanese.
Korea rebuts
① The three islands are examples among Korea's 3,000-plus islands — omission does not make an island Japanese. ② Nothing in the treaty reverses SCAPIN 677. ③ The deletion reflects American avoidance of the question; UK, Australian and NZ drafts diverged — a multilateral treaty cannot be construed by one state's intent.
original published documented MOFA Japan commentaryNAHF commentary

The Rusk letter 1951.8.10

The State Department's reply refusing Korea's request to add "Dokdo and Parangdo" to Article 2(a): the rocks were "never treated as part of Korea" and had been under Shimane's Oki branch office "since about 1905."

…this normally uninhabited rock formation was according to our information never treated as part of Korea…
Meaning
The decisive link in Japan's treaty argument: the drafting power explicitly rejected Korea's request.
Korea rebuts
① A confidential note delivered only to Korea (published c. 1978) — no basis for interpreting a 48-party treaty. ② "Our information" appears to derive from Japanese-supplied materials, ignoring Edict 41 and the 1877 directive. ③ Dulles cabled in 1953 that the Rusk position was "only one of many signatories'" views; internal US papers settled on non-intervention, and the US has been formally neutral since.
original published documented Wikisource full text + scansMOFA Japan posting

Three ICJ referral proposals 1954 · 1962 · 2012

Japan proposed referring the dispute to the ICJ three times; Korea refused each time. ICJ jurisdiction requires both parties' consent — Korea has never accepted compulsory jurisdiction, so a unilateral Japanese filing cannot open a case.

Meaning
Japan's framing: "we seek resolution by law and dialogue; Korea avoids the courtroom."
Korea rebuts
"No territorial dispute exists, so there is nothing to adjudicate" (Foreign Minister Byun's 1954 statement called the proposal "a false attempt dressed in judicial procedure"). Being labeled 'disputed' is itself the effect Japan seeks. Note the structural symmetry: Japan refuses ICJ referral over the Senkakus it controls, with the same 'no dispute' logic — possessors have only downside at court.
original published documented Cabinet SecretariatOpinio Juris on the symmetry

Post-1905 administration · Takeshima Day 1905~ · 2005~

After incorporation, Shimane licensed sea-lion hunting and collected land-use fees. In 2005 the prefectural assembly created "Takeshima Day" (Feb 22), marked annually (with a central-government vice-minister attending since 2013).

Meaning
The claim of "peaceful and continuous display of state authority without protest" — built to the standard of ICJ case law (Palmas, Minquiers).
Korea rebuts
① 'No protest' reflects Korea's loss of diplomatic capacity after 1905 — impossibility of protest cannot count as acquiescence (the 1906 Sim Heung-taek report records denial). ② Acts built on an unlawful taking create no title (ex injuria jus non oritur). ③ The sea-lion fishery exterminated the Dokdo sea lion — exploitation, not stewardship.
copies published documented Shimane Takeshima pages

Same text, two readings 3 battlegrounds

Where the two sides read the very same document in opposite ways — pick a reading and watch the highlighter move. The initial selection is random.

① Is 'Usando' Dokdo? — Sejong geography (1454)

于山武陵二島 在縣正東海中 二島相去不遠 風日淸明 則可望見

"A distinct, distant island visible only on clear days" — from Ulleungdo, the only such island is Dokdo. The 1770 Dongguk Munheon Bigo states "Usan is what Japan calls Matsushima (Dokdo)."

'Usan' is an alias of Ulleungdo or a phantom island — there is no proof the two islands are Ulleungdo and Dokdo, and the 1770 gloss merely recycles An Yong-bok's unreliable testimony.

于山島 鬱陵島
The 1531 atlas layout — Usando drawn west of Ulleungdo at similar size. 🇯🇵 the "Usando ≠ Dokdo" argument
Ulleungdo Dokdo 87.4 km
Actual geography — Dokdo lies 87.4 km southeast of Ulleungdo, visible on clear days. 🇰🇷 the "can be seen" argument

② Is 'Seokdo' Dokdo? — Edict No. 41 (1900)

區域은 鬱陵全島와 竹島 石島를 管轄할 事

Stone island → dialectal dok island → 石島 by meaning, 獨島 by sound — the gazette transcribing the settlers' name. The 1906 report "Dokdo, under this county" completes the chain.

If Seokdo is Dokdo, why not write 'Dokdo' or 'Usando'? Some identify Seokdo as Gwaneumdo — and even granting it, there is no evidence of control around 1900.

③ The treaty's silence — Art. 2(a) (1951)

Japan… renounces all right, title and claim to Korea, including the islands of Quelpart, Port Hamilton and Dagelet.

'including' introduces examples — you cannot list three thousand islands. Omission does not mean Japanese ownership, and nothing reverses SCAPIN 677.

The three named islands do not include Takeshima — and Korea's request to add it was expressly refused in the Rusk letter. The treaty therefore affirms it as Japanese.

Third-party views neutral sources

Finding: truly neutral sources are rare. The accessible English-language legal literature converges on Korea's side, but each work carries caveats — the safest neutral construction is ① institutional explainers ② the naming record ③ the US 'no position' stance ④ both governments' materials presented side by side.

The US official position neutral since the 1950s

The decades-long line: the US "takes no position" on sovereignty over the Liancourt Rocks. In 2008 the Board on Geographic Names briefly reclassified the rocks as 'Undesignated Sovereignty' — reversed within a week on President Bush's instruction: "this is an issue for Korea and Japan to resolve" (KBS interview).

Value / limits
Both the 1951 Rusk letter (cited by Japan) and the 1946 SCAPINs (cited by Korea) are American documents — the US archive feeds both arguments at once. The 2008 reversal shows the neutrality is politically managed.
original published asserted by a government Bush KBS interview (White House archive)

The three-way map labels 2012 policy

Google Maps labels the island 'Dokdo' in the Korean edition, '竹島' in the Japanese, and 'Liancourt Rocks' internationally (Oct 2012 policy change; Korea's foreign ministry protested). The CIA World Factbook likewise uses 'Liancourt Rocks' throughout.

Value / limits
The factual basis for 'one island, three maps.' The name traces to the French whaler Le Liancourt, which sighted the rocks in 1849 — a naming history neither side disputes.
copies published documented Korea Herald 2012Liancourt Rocks (Wikipedia)

The legal literature 2005 · 2007

Van Dyke (2007): applying ICJ case law, concludes Korea's title is "substantially stronger" (estoppel from the 1905 terra-nullius declaration, the taint of annexation, impossibility of protest, contiguity, a half-century of control) — while judging the rocks an Art. 121(3) 'rock' generating no EEZ. Fern (2005): using the neutral name throughout, finds Korea has demonstrated more "affirmative acts of sovereignty" — and that neither side has any incentive to adjudicate.

Value / limits
Van Dyke acknowledges funding from a Korea-America research center, so cite with care; Fern is a student journal. English-language work concluding for Japan comes mainly from government or Sasakawa-affiliated channels — no independent journal publication was verified in this survey.

How ICJ jurisdiction works institutional

The Court's contentious jurisdiction rests entirely on state consent (special agreement / treaty clause / optional-clause declaration / forum prorogatum). Possessor states avoiding court is not unique to this island — Japan does the same over the Senkakus, Britain over the Falklands.

Value / limits
The Court's own institutional explainer — the most neutral source in this dossier. It does not mention the island.
original published ICJ — basis of jurisdiction

The two official sites for parallel reading

🇰🇷 MOFA Korea's Dokdo site: 11 languages — "clearly Korean territory historically, geographically and under international law; no dispute exists." 🇯🇵 MOFA Japan's Takeshima site: 12 languages — "inherent Japanese territory in light of historical facts and international law; Korea occupies it illegally" + the 28-page '10 Points' PDF.

Value / limits
The wellsprings of both official narratives. Note: mofa.go.jp returns 403 to automated access; normal browsers load fine.
Which story do you find more convincing?
Past records are archived forever
counting — split shown once enough votes are in
The frozen record — never edited, never deleted

Method and limits. Compiled 2026-08-24 via three parallel research tracks (Korean grounds / Japanese grounds / third parties). Every URL was verified live; Japanese government sites block automated access, so content was confirmed via browser and archives. Neither side disputes that these documents exist — the entire quarrel is interpretive. This dossier therefore pairs every document with a rebuttal, and leaves judgment to the reader. Three of the twelve carry a further answer to that rebuttal as well, attributed and visible as a change of speaker; all three are on Korean cards and no Japanese card carries one. A rebuttal that is answered and a rebuttal that stands are not the same thing, and the reader should be able to see which is which. Badges: original published = images or full text publicly available from an institution; copies published = original lost or not extant, transmitted in copies.

Editorial principles. ① Name order follows the language edition: Dokdo leads in Korean, Takeshima in Japanese; third-language editions prefer the neutral name (Liancourt Rocks). Where no neutral name exists, the controller's name leads, labeled as such. ② The order of the two narrative sections is randomized on every visit. ③ Every dispute is labeled with its de facto controller, and every source carries the other side's rebuttal.