JP-J-01: Three Accounts of the Comfort Women Issue and the 2015 Agreement (1932–1945; 1991–2026)

Status: DRAFTWords: 10,158

1. Key Takeaways

The wartime "comfort women" (ianfu) system — the network of military-organised brothels ("comfort stations", ianjo) established and operated across the Japanese empire and its occupied territories between approximately 1932 (the first documented Shanghai-area stations following the First Shanghai Incident) and 1945 — is among the most intensively contested subjects in East Asian historical-memory politics, and this document deliberately presents it through three distinct interpretive accounts rather than a single composite narrative: the survivor-centred/transnational-justice account, the Japanese conservative/revisionist account, and the diplomatic-realist/reconciliation-studies account. The three accounts share a factual core — comfort stations existed across the empire; the Japanese military was involved in their establishment, regulation, and operation; women from Korea, China, the Philippines, Indonesia (including Dutch internees), Taiwan, Burma, and Japan itself were present in them; many of those women suffered grievously — and diverge fundamentally on the definition of coercion, the scale of the system, the form of responsibility owed, and the possibility of "final" resolution. The corpus declares no winner among the three.
The numerical question is itself an artefact of the contestation and every figure must be carried with its attribution. Estimates of the total number of women in the system range from approximately 20,000 (the low-end figure associated with the Japanese historian Hata Ikuhiko's revised calculations) through 50,000–200,000 (the range cited in the 1996 United Nations Coomaraswamy Report) to 200,000-plus (the figure used by the Korean Council and much of the transnational advocacy network, sometimes extended to 400,000 in Chinese-source claims) [TBD-VERIFY: each figure is an estimate resting on contested extrapolation methods — soldier-to-comfort-woman ratios, station counts, replacement-rate assumptions — and no documentary census of the system survives or likely ever existed; the corpus treats the number as genuinely unknown within a wide band and attributes every specific figure to its source]. The compositional question — what proportion of the women were Korean, Japanese, Chinese, or other — is similarly contested, with Korean-majority claims standard in the advocacy literature and Japanese-majority-among-licensed-prostitutes claims standard in the revisionist literature [TBD-VERIFY: compositional proportions by nationality].
The issue's modern political career began not in 1945 but on 14 August 1991, when Kim Hak-sun (1924–1997), a Korean survivor, gave the first public, named, on-the-record testimony by a former comfort woman at a Seoul press conference organised with the Korean Council for the Women Drafted for Military Sexual Slavery by Japan (Chongdaehyop, founded November 1990), followed by her December 1991 suit against the Japanese government in the Tokyo District Court. The four-and-a-half decades of near-silence between 1945 and 1991 — produced by survivor stigma, by the patriarchal social orders of both post-war Japan and post-war Korea, by the Cold War security architecture that subordinated historical claims to anti-communist alignment, and by the 1965 Treaty framework's "completely and finally" settlement language — is itself a central analytical datum: the issue entered international politics only when Korean democratisation (post-1987), the global women's-rights movement, and the end of the Cold War converged to make survivor testimony speakable and politically consequential.
The Japanese documentary record was transformed in January 1992 when the historian Yoshimi Yoshiaki (Chuo University) published documents he had located in the library of the National Institute for Defense Studies — including a 1938 directive of the North China Area Army and Ministry of War documents — establishing that the Japanese military had been directly involved in the establishment and management of the comfort stations, contradicting the then-standing Japanese-government position that the stations had been purely private commercial operations. The Yoshimi discoveries, published in the Asahi Shimbun on 11 January 1992 days before Prime Minister Miyazawa Kiichi's Seoul visit, forced the first Japanese-government acknowledgement (the Kato Koichi statement of July 1992) and conditioned the 4 August 1993 Kono Statement, issued by Chief Cabinet Secretary Kono Yohei, which acknowledged that the Japanese military was "directly or indirectly" involved in the establishment and management of the comfort stations, that recruitment was conducted "mainly by private recruiters" but "in many cases" against the women's will, "through coaxing, coercion, etc.", and that "at times, administrative/military personnel directly took part in the recruitments". The Kono Statement has remained the formal Japanese-government position through every subsequent cabinet, including both Abe cabinets — even as its meaning has been contested and qualified from within.
A parallel and corrosive evidentiary subplot is the Yoshida Seiji fabrication. Yoshida (1913–2000), a former wartime labour-mobilisation official, claimed from 1977 and in his 1983 book My War Crimes (Watashi no Sensō Hanzai) to have personally led "slave hunts" (ianfu-gari) forcibly seizing Korean women on Jeju Island; his testimony was widely reported, including in at least sixteen Asahi Shimbun articles, and was cited in early international documentation including the 1996 Coomaraswamy Report. Scholarly doubt was raised as early as 1992 by Hata Ikuhiko's Jeju field investigation (building on the Jeju local newspaper Cheju Shinmun's 1989 checking of Yoshida's claims), and on 5 August 2014 the Asahi Shimbun formally retracted all articles based on Yoshida's testimony as fabricated. The retraction's significance is read oppositely by the accounts: the revisionist account treats it as vindication of the claim that the "forcible-abduction" narrative was built on fraud; the survivor-centred account responds that the case for the system's coercive character never rested on Yoshida — it rests on survivor testimony across multiple countries, on the Semarang case's judicially established facts, and on the Yoshimi documents — and that the retraction's weaponisation illustrates the revisionist strategy of attacking the weakest evidentiary plank to discredit the whole.
The remedial sequence has run through four major instruments, each contested. First, the Asian Women's Fund (1995–2007), established under the Murayama cabinet as a private-donation "atonement money" (tsugunaikin) mechanism (¥2 million per recipient, accompanied by signed letters of apology from successive prime ministers and government-funded medical-welfare support), which was rejected by a majority of Korean survivors and by the Korean Council on the ground that its private character evaded state legal responsibility [TBD-VERIFY: precise recipient counts — commonly reported as 285 women in Korea, the Philippines, and Taiwan accepting atonement money, with additional programmes in the Netherlands and Indonesia]. Second, the 2007 first-Abe-cabinet position — the 16 March 2007 cabinet-approved answer stating that no documents had been found directly showing "coercion in the narrow sense" (forcible abduction by officials), which triggered US House of Representatives Resolution 121 (30 July 2007) calling on Japan to formally acknowledge and apologise. Third, the 28 December 2015 Agreement. Fourth, the post-2015 unravelling and the post-2023 quieting under the Yoon-period bilateral reset documented in JP-F-03.
The 28 December 2015 Agreement — announced in Seoul through parallel press statements by Foreign Ministers Kishida Fumio and Yun Byung-se, following Director-General-level consultations across 2014–2015 and a Park Geun-hye-Abe Shinzo summit channel facilitated by sustained Obama-administration alliance-management pressure — comprised: Abe's expression, conveyed by Kishida, of "his most sincere apologies and remorse to all the women who underwent immeasurable and painful experiences"; the acknowledgement that the issue occurred "with an involvement of the Japanese military authorities" and that the Government of Japan "is painfully aware of responsibilities"; a Japanese-government budget contribution of ¥1 billion to a Korean-government-established foundation (the Reconciliation and Healing Foundation, established July 2016); and the mutual confirmation that the issue would be "resolved finally and irreversibly" provided the agreement was faithfully implemented, with both governments to refrain from criticising each other on the issue in the international community. The agreement also contained the unwritten-understandings dispute in embryo: the Korean side "acknowledged" Japanese concern over the comfort-woman statue (Pyeonghwabi, the Statue of Peace) before the Japanese Embassy in Seoul and undertook to "strive to solve this issue in an appropriate manner" — language the Japanese side treated as a relocation expectation and the Korean side treated as a non-commitment.
The unravelling was rapid and structurally instructive. The agreement was concluded without the survivors' participation; the Korean Council and a substantial bloc of survivors rejected it immediately; the December 2016 placement of a second statue before the Japanese Consulate-General in Busan triggered Japan's January 2017 temporary recall of its ambassador; the Moon Jae-in administration's December 2017 task-force review found the negotiation procedurally flawed and partially secret; the Reconciliation and Healing Foundation was dissolved (announced November 2018, registration cancelled July 2019) with [TBD-VERIFY: roughly 34–36 of the 47 survivors alive at the agreement date reported as having accepted the foundation's payments before dissolution]; and the January 2021 Seoul Central District Court ruling ordering the Japanese state to compensate twelve plaintiffs — rejecting sovereign immunity for jus cogens violations — was followed in April 2021 by a different bench of the same court dismissing a parallel suit on sovereign-immunity grounds, leaving Korean jurisprudence itself split [TBD-VERIFY: the November 2023 Seoul High Court appellate ruling reversing the April 2021 dismissal and finding jurisdiction]. The Japanese position throughout has been that the 2015 agreement remains in force and that the Korean court proceedings violate international law.
By the mid-2020s the issue had entered a demographic endgame and a geopolitical quieting. The registered Korean survivor cohort — 240 women registered with the Korean government since 1993 — had declined to single digits [TBD-VERIFY: South Korean Ministry of Gender Equality and Family figures indicating fewer than ten registered survivors alive by 2025–2026, with deaths across 2023–2025 reducing the cohort from the low teens]; the Yoon Suk-yeol administration (2022–2025) deliberately de-emphasised the issue within its broader third-party-reimbursement approach to the forced-labour portfolio and its Camp David trilateral alignment; and the Lee Jae-myung administration (June 2025–) has, in its first year, maintained a "two-track" framing that keeps historical-memory grievances rhetorically alive while not re-escalating the comfort-women file specifically [TBD-VERIFY: the Lee-administration's specific posture on the 2015 agreement's status as of mid-2026 — the available record through the JP-F-03 coverage window indicates national-interest-centred review language on the forced-labour settlement without a parallel formal reopening of the comfort-women agreement]. The diplomatic-realist account reads this quieting as confirmation that geopolitics, not justice, drives the issue's cycles; the survivor-centred account reads it as the predictable consequence of resolving a justice claim by demographic attrition; the revisionist account reads it as the issue's instrumental character finally exhausting itself.
The 2015 agreement functions in this document as a test object: a single, well-documented diplomatic artefact that each account reads through its own logic and that therefore exposes the accounts' deep structures. The survivor-centred account reads it as procedurally illegitimate (negotiated over the survivors' heads) and substantively insufficient (no admission of legal responsibility, no Diet resolution, no individual state compensation as of right). The revisionist account reads it as Japan's third and final over-payment on a claim already settled in 1965 and re-settled in 1995, whose Korean repudiation proves the "moving goalposts" thesis. The diplomatic-realist account reads it as a technically sound, mutually painful compromise — the most explicit Japanese acknowledgement-plus-state-funds package ever assembled — that failed because neither government invested the domestic political capital to ratify it socially: Park concluded it without survivor buy-in months before her own political collapse, and Abe undercut it domestically by assuring his base that the apology marked no new admission. The three readings are not reconcilable; their juxtaposition is the document's method.

2. The Factual Spine: The System and the Issue's Career (1932–2026)

Before the three accounts are presented, this section sets out the chronological record at the level of detail on which the accounts themselves substantially agree, flagging the points where the factual record itself is contested.

The wartime system, 1932–1945. The first documented military comfort stations were established in the Shanghai area in 1932, in the aftermath of the First Shanghai Incident, reportedly at the initiative of Lieutenant-General Okamura Yasuji, with the stated rationales of preventing rapes of local civilians by Japanese troops (after the international damage done by documented atrocities), controlling venereal disease through medically regulated facilities, and preventing espionage through unregulated contact between soldiers and local women. The system expanded massively after the outbreak of full-scale war with China in July 1937 and particularly after the Nanjing atrocities of December 1937–January 1938, when the army leadership concluded that regulated facilities were preferable to uncontrolled assaults. By the Pacific War's end, comfort stations had operated across China, Korea, Manchuria, Taiwan, the Philippines, Burma, Malaya, Singapore, the Dutch East Indies, French Indochina, the Pacific islands, and Japan-proximate territories including Okinawa [TBD-VERIFY: total station counts — Yoshimi Yoshiaki's work documents stations in every major theatre, but no comprehensive census of stations exists; estimates of station numbers range widely].

The women in the system came from across the empire and its conquests. Korean women — recruited under conditions ranging from outright deception (promises of factory or nursing work) through debt bondage and sale by impoverished families to, in contested degrees, direct coercion — formed a large component, reflecting Korea's colonial status since 1910 and the recruitment networks (largely operated by private brokers, many themselves Korean, under military licensing and logistical support) that channelled women to the front [TBD-VERIFY: the proportion of Korean women and the distribution of recruitment modes are both centrally contested between the accounts]. Chinese women were taken in occupied areas, frequently — on the documentation assembled by Chinese researchers and in cases adjudicated in post-war Chinese and Japanese court proceedings — by direct violence in war zones. Filipina women's testimonies, collected from 1992 by the Task Force on Filipino Comfort Women and Lila Pilipina, include numerous accounts of seizure by soldiers and confinement in garrison buildings. Japanese women, largely drawn from the existing licensed-prostitution industry, were present throughout and may have predominated in the system's early years [TBD-VERIFY: compositional claim]. Taiwanese, Burmese, Malayan, Indonesian, and Timorese women appear across the documentation.

The Semarang case. The single most juridically solid episode is the Semarang (Java) case of early 1944: Japanese military officers removed approximately thirty-five Dutch women and girls from internment camps in the Semarang area and forced them into comfort stations, an episode terminated after roughly two months when senior officers in Tokyo learned of it. A 1948 Batavia (Jakarta) Temporary Court-Martial convened by the returning Dutch colonial authorities convicted eleven Japanese military personnel and one civilian; one officer, Major Okada Keiji, was sentenced to death and executed [TBD-VERIFY: precise sentence distribution among the twelve convicted]. The Semarang convictions matter analytically because they constitute a contemporaneous judicial finding of forcible enslavement by military personnel — a finding the Japanese government has never disputed and which the 1994 Dutch government study (the report to the Dutch parliament on the forced prostitution of Dutch women in the Dutch East Indies) reinforced, estimating that of the 200–300 European women working in comfort stations in the Indies, some sixty-five were "most certainly forced into prostitution" [TBD-VERIFY: Dutch report figures]. One survivor of the Semarang episode, Jan Ruff-O'Herne (1923–2019), became from 1992 one of the issue's most internationally prominent witnesses, testifying to the US House of Representatives in 2007.

The silence, 1945–1991. Between Japan's surrender and 1991 the comfort-women issue had almost no public political existence. The Tokyo Tribunal (IMTFE, 1946–1948) heard evidence of forced prostitution in certain theatres but did not treat the system as a distinct charge; the Batavia trials were the exception, prosecuting only the Dutch-victim cases. The 1951 San Francisco Treaty and the bilateral settlements that followed — above all the June 1965 Treaty on Basic Relations between Japan and the Republic of Korea and its accompanying Claims Settlement Agreement, under which Japan provided US$300 million in grants and US$200 million in loans and both states agreed that claims between them and their nationals were "settled completely and finally" — were negotiated without any reference to the comfort women, whose existence was publicly unspoken in both countries. In Korea, survivors faced intense stigma in a patriarchal social order; in Japan, the subject appeared mainly in soldier memoirs and in Senda Kako's 1973 book Jūgun Ianfu (Military Comfort Women), which named the phenomenon for a mass readership but produced no state response. The 1970s–1980s Korean women's movement and Japanese feminist scholarship slowly built the infrastructure of attention; Yun Chung-ok's research and her January 1990 Hankyoreh series "In the Footsteps of the Voluntary Service Corps" was a proximate catalyst, and the Korean Council for the Women Drafted for Military Sexual Slavery by Japan (Chongdaehyop) was founded on 16 November 1990 as a coalition of Korean women's organisations.

Kim Hak-sun and the eruption, 1991–1993. On 14 August 1991 Kim Hak-sun testified publicly in Seoul; on 6 December 1991 she and two other survivors (initially anonymous) joined a class-action suit in the Tokyo District Court. In January 1992 Yoshimi's documents were published; Prime Minister Miyazawa apologised in general terms during his January 1992 Seoul visit; Chief Cabinet Secretary Kato Koichi issued a first acknowledgement-statement in July 1992; and on 4 August 1993, following a Japanese government study that included interviews with sixteen Korean survivors, Chief Cabinet Secretary Kono Yohei issued the Kono Statement. The statement's carefully layered language — military involvement "directly or indirectly" in establishment and management; recruitment "conducted mainly by private recruiters who acted in response to the request of the military", but "in many cases" women "were recruited against their own will, through coaxing, coercion, etc."; "at times, administrative/military personnel directly took part in the recruitments"; and the women "lived in misery at comfort stations under a coercive atmosphere" — was the product of intense negotiation within the Japanese government and, as a 2014 Japanese government review confirmed, of textual consultation with Seoul. It included an apology and remorse and a commitment to face history; it did not concede legal liability beyond the 1965 framework.

The Yoshida Seiji fabrication and its long shadow. Yoshida Seiji's claims to have led armed "comfort women hunts" on Jeju Island, published from 1977 and in his 1983 book, were the most graphic first-person Japanese-perpetrator testimony in circulation, reported repeatedly by the Asahi Shimbun and other outlets and cited in international documentation. The Jeju newspaper Cheju Shinmun found no local corroboration as early as 1989; Hata Ikuhiko's 1992 field investigation on Jeju found none either, and Yoshida later declined to defend the specifics, at one point telling an interviewer that a book may mix fact and fiction [TBD-VERIFY: Yoshida's exact 1996 statement to Shūkan Shinchō is variously rendered]. On 5 August 2014 the Asahi retracted at least sixteen articles resting on his testimony; the affair, combined with a separate Asahi retraction controversy over Fukushima reporting, produced a major credibility crisis for the paper, a third-party committee review, and the resignation of Asahi president Kimura Tadakazu in December 2014. The Kono Statement itself, the 2014 Japanese government review found, had not relied on Yoshida's testimony [TBD-VERIFY: the review's treatment of Yoshida is commonly so summarised; the review text should be checked directly].

The Asian Women's Fund, 1995–2007. Established in July 1995 under the Murayama Tomiichi coalition cabinet, the Asian Women's Fund (Josei no tame no Ajia Heiwa Kokumin Kikin) offered each accepted applicant ¥2 million in "atonement money" raised from private donations, government-funded medical and welfare support of ¥1.2–3 million depending on country, and a signed letter of apology from the sitting prime minister expressing "apologies and remorse" for the actions of the Imperial military. The Fund operated programmes in Korea, the Philippines, Taiwan, the Netherlands, and Indonesia (where the money funded elder-care facilities rather than individual payments, by agreement with Jakarta). The Korean Council campaigned against acceptance, characterising the Fund as a private-charity evasion of state legal responsibility, and the Kim Young-sam and Kim Dae-jung governments ultimately provided counterpart payments to survivors who declined the Fund. [TBD-VERIFY: commonly cited figures — 285 women in Korea, the Philippines, and Taiwan received atonement money; 79 Dutch women received project support — rest on the Fund's own final reporting.] The Fund was dissolved in March 2007, regarded by its own architects, including the historian Onuma Yasuaki, as a partial failure of execution and communication rather than of conception.

2007: the Abe cabinet statement and House Resolution 121. On 16 March 2007 the first Abe cabinet approved a written answer to a Diet question stating that the materials the government had located did not directly show "coercion in the narrow sense" — forcible abduction by military or officials — in recruitment. The statement, alongside Abe's earlier associations with the revisionist Diet caucus on historical education, internationalised the issue: on 30 July 2007 the US House of Representatives passed House Resolution 121 (sponsored by Representative Mike Honda) calling on Japan to "formally acknowledge, apologize, and accept historical responsibility in a clear and unequivocal manner"; parallel resolutions followed in the Dutch, Canadian, and European parliaments. Abe stated during his April 2007 Washington visit that he had sympathy for the former comfort women and apologised "as prime minister" in general terms, while his government maintained the narrow-coercion documentary position.

2011–2015: the Constitutional Court ruling, statue diplomacy, and the road to the agreement. On 30 August 2011 the Constitutional Court of Korea ruled that the Korean government's failure to make active diplomatic efforts to resolve the survivors' compensation claims — given the interpretive dispute over whether the 1965 agreement covered them — was unconstitutional, converting the issue from a civil-society demand into a Korean state obligation. On 14 December 2011, at the 1,000th weekly Wednesday Demonstration outside the Japanese Embassy in Seoul (a protest series running since 8 January 1992), the Korean Council installed the Statue of Peace (Pyeonghwabi), a seated bronze girl by the sculptors Kim Seo-kyung and Kim Eun-sung, facing the embassy. Statue placements multiplied: Glendale, California (July 2013, upheld in US federal litigation); Busan, before the Japanese Consulate-General (December 2016); San Francisco (a column-of-strength memorial, September 2017, which prompted Osaka to terminate its sister-city relationship with San Francisco in October 2018); Berlin-Mitte (September 2020, subject to repeated removal orders and extensions thereafter [TBD-VERIFY: the Berlin statue's status as of 2024–2026 — the district moved toward removal/relocation in 2024 amid litigation]); and others in Australia, Canada, and the United States [TBD-VERIFY: current statuses of individual overseas installations]. Park Geun-hye, taking office in February 2013, made the comfort-women issue a precondition for a bilateral summit with Abe; twelve rounds of director-general-level consultations ran from April 2014; the Obama administration, alarmed by the Japan-Korea freeze amid North Korean and Chinese pressures, pushed both sides toward settlement, including through the March 2014 Hague trilateral meeting; and Abe and Park held their first bilateral summit on 2 November 2015 in Seoul, instructing negotiators to accelerate.

The 28 December 2015 Agreement and its unravelling, 2015–2019. The agreement's content is given in the Key Takeaways above. Implementation began on the Korean side with the July 2016 establishment of the Reconciliation and Healing Foundation under chairwoman Kim Tae-hyeon and Japan's August–September 2016 transfer of ¥1 billion. The foundation offered 100 million won to survivors alive at the agreement date and 20 million won to families of the deceased; of the forty-seven registered survivors alive on 28 December 2015, those accepting payments before the foundation's dissolution are commonly reported in the mid-thirties [TBD-VERIFY: 34, 35, and 36 are all reported in different sources]. But the agreement's social ratification failed in Korea: the surviving leadership of the survivor movement rejected it; the Busan consulate statue appeared in December 2016 and Japan recalled Ambassador Nagamine Yasumasa for nearly three months from January 2017; Park was impeached (December 2016) and removed (March 2017); and Moon Jae-in, elected May 2017 on a platform critical of the deal, commissioned a Foreign Ministry task-force review which reported on 27 December 2017 that the negotiation had been victim-excluding and had included unannounced understandings (on the statue, on third-country memorials, and on the term "sexual slavery"). Moon declared the agreement "seriously flawed" but did not formally abrogate it or return the ¥1 billion; the foundation's dissolution was announced on 21 November 2018 and its registration cancelled in July 2019.

2020–2026: the Yoon Mee-hyang scandal, the sovereign-immunity split, and the demographic endgame. In May 2020 the survivor Lee Yong-soo publicly accused Yoon Mee-hyang — the long-time Korean Council leader, newly elected to the National Assembly — of exploiting survivors and misusing donations; Yoon was indicted on accounting and embezzlement-related charges, largely acquitted at first instance in February 2023, and received a suspended sentence on a partial conviction at appeal in 2023 [TBD-VERIFY: final disposition at the Supreme Court of Korea]. On 8 January 2021 the Seoul Central District Court (Judge Kim Jeong-gon's bench) ordered Japan to pay 100 million won each to twelve plaintiffs, holding that sovereign immunity does not shield grave human-rights violations; Japan, which had refused to appear, rejected the ruling; on 21 April 2021 a different bench of the same court dismissed a second, larger suit on sovereign-immunity grounds; and in November 2023 the Seoul High Court reversed that dismissal and entered judgment against Japan, which again refused to recognise the proceedings [TBD-VERIFY: enforcement steps, if any, against Japanese state assets remain unresolved]. The Yoon Suk-yeol administration (May 2022–) subordinated the issue to its alliance-first reset documented in JP-F-03 — the March 2023 third-party-reimbursement formula on forced labour, the resumed shuttle summitry, and the August 2023 Camp David trilateral — and stated that the 2015 agreement remained the two governments' official position. The registered survivor cohort declined from 240 registrants (cumulative since 1993) to single digits by 2025–2026 [TBD-VERIFY: exact count as of mid-2026]. The Lee Jae-myung administration (June 2025–) has maintained two-track language without formally reopening the 2015 agreement [TBD-VERIFY: against the JP-F-03 coverage line and subsequent developments].

3. Account 1: The Survivor-Centred / Transnational Justice Account

Evidentiary foundations. The first account is grounded in four bodies of material. First, survivor testimony itself: the recorded testimonies of hundreds of women across Korea (the Korean Council and the Korea Chongshindae Research Institute published six volumes of collected testimonies from 1993), China, the Philippines, Indonesia, Taiwan, East Timor, Malaysia, and the Netherlands, exhibiting — on this account's reading — a cross-national consistency of core experience (deceptive or coerced recruitment, confinement, repeated rape, violence, disease, abandonment at war's end) that no coordination could have manufactured. Second, the documentary record assembled by Yoshimi Yoshiaki and the Center for Research and Documentation on Japan's War Responsibility (JWRC, founded 1993): military directives on station establishment, regulations governing operations, transport authorisations for women crossing military lines, and venereal-disease records, collected in Yoshimi's Jūgun Ianfu (1995; English: Comfort Women: Sexual Slavery in the Japanese Military During World War II, 2000). Third, the international legal architecture: the January 1996 report of UN Special Rapporteur on Violence against Women Radhika Coomaraswamy, which characterised the system as "military sexual slavery" and recommended state compensation, prosecution, and apology; the June 1998 report of Special Rapporteur Gay McDougall for the Sub-Commission on Prevention of Discrimination, which analysed the stations as "rape centres" and the system as a crime against humanity under the law of the time; and the repeated findings of UN treaty bodies (CEDAW, the Human Rights Committee, the Committee against Torture) urging Japan toward fuller remedy. Fourth, the Women's International War Crimes Tribunal on Japan's Military Sexual Slavery (Tokyo, December 2000) — a civil-society "people's tribunal" with no legal force, presided over by Gabrielle Kirk McDonald, former president of the ICTY — which heard sixty-four survivors and rendered a December 2001 judgment finding Emperor Hirohito and the Japanese state "guilty" of crimes against humanity. The Semarang convictions and the 1994 Dutch government report anchor the account's claim that forcible enslavement by military personnel is not an inference but an adjudicated fact in at least one theatre.

Core claims. The account's claims can be stated in five propositions. (1) The system was military sexual slavery: the controlling fact is not the recruitment moment but the condition of servitude — confinement, the inability to refuse, the absence of meaningful consent or exit — and on that definition the system was enslavement regardless of whether a given woman was abducted at bayonet-point, deceived by a broker, or sold by her family, because the military created the demand, licensed the supply chain, regulated the stations, and held the women within them. (2) Coercion took many forms: the revisionist demand for documents proving forcible abduction by uniformed officials is, on this account, a category error that defines coercion down to its rarest and most documentable form, ignoring deception, debt bondage, intimidation in a colonial setting where refusal of authority carried danger, and the structural coercion of girls aged in some cases under fifteen [TBD-VERIFY: age distributions rest on testimony and partial records]. (3) State legal responsibility remains undischarged: the 1965 treaty settled economic claims between states without the comfort-women issue ever being raised; grave violations of human dignity generate obligations — acknowledgement of legal (not merely moral) responsibility, compensation as of right, full disclosure of records, accurate education, and guarantees of non-repetition — that lump-sum economic diplomacy cannot extinguish, a position the 2011 Korean Constitutional Court ruling and the January 2021 Seoul judgment both partially vindicate. (4) The 2015 agreement failed for a discoverable reason: it was negotiated over the survivors' heads, announced to them rather than with them, and conditioned on silencing ("refrain from accusing or criticizing each other... in the international community") — and a settlement of a dignity claim that the holders of the claim reject is not a settlement; the principle "nothing about us without us" is not sentimentality but the operational lesson of the Asian Women's Fund's failure, repeated. (5) "Final and irreversible" cannot be imposed on victims: governments may bind each other; they cannot bind the conscience of survivors or the historiography of nations, and the attempt to do so converted a partial apology into an instrument that survivors experienced as a second silencing.

The account's treatment of the difficult evidence. The account does not depend on Yoshida Seiji and, in its more rigorous registers (Yoshimi himself, the JWRC), discarded him early; it acknowledges that the Asahi retraction damaged public trust but insists the documentary and testimonial record stands independently. It acknowledges that brokers, many Korean, ran much of the recruitment — and answers that licensing and using a coercive supply chain is the responsibility of the principal, not only the agent. It acknowledges the 200,000 figure is an advocacy estimate and that the true number is unknown — while observing that the destruction of records at the war's end, ordered across the Japanese state in August 1945, is itself part of why the number cannot be known.

Weakest points. Stated within the corpus's discipline of arming each account's critics: (1) the account's public-facing wing has at times treated the most expansive numerical claims (200,000+, "mostly teenage girls forcibly taken") as settled fact when they are estimates resting on weak extrapolation, creating exposure that revisionists exploit; (2) testimony collected decades after the events, in politically charged settings, sometimes evolved across retellings — a normal feature of trauma memory, but one the account's institutional wing has been reluctant to address methodologically, and the early citation of Yoshida by the movement and by the Coomaraswamy Report remains a genuine blemish; (3) the Yoon Mee-hyang scandal of 2020 — with the survivor Lee Yong-soo herself charging that the Korean Council had instrumentalised survivors, kept them poor, and centred the organisation rather than the women — damaged the claim that the advocacy infrastructure straightforwardly spoke for survivors, and exposed the awkward fact that some survivors wished to accept Japanese money (in 1995–97 and again in 2016–17) and faced movement pressure not to; (4) the account has no operational answer to the finality problem: if "final and irreversible" cannot be imposed, what settlement could ever conclude the issue, and is the implicit answer — none, until the last survivor's terms are met — compatible with the existence of inter-state dispute resolution at all? Critics including reconciliation scholars note that the account's standards rose over time (from apology, to apology-plus-compensation, to legal responsibility, to legislative enactment), which is defensible as deepening understanding but is also functionally indistinguishable, to the Japanese side, from goalposts that move.

4. Account 2: The Japanese Conservative / Revisionist Account

Evidentiary foundations. The second account — articulated across a spectrum running from rigorous empiricist historians through the Sankei Shimbun and Yomiuri editorial lines to polemical deniers, and the corpus distinguishes these registers — is grounded in: Hata Ikuhiko's scholarship (Ianfu to senjō no sei / Comfort Women and Sex in the Battle Zone, 1999), which assembled the licensed-prostitution context, the broker economics, the wage records, and the demographic recalculations producing his ~20,000 estimate; the contemporaneous-documents debate — the observation that no order directing the forcible abduction of women has been found in Japanese archives, and that some recovered documents (such as the Home Ministry and army directives of 1938 instructing care in recruitment practices to avoid kidnapping-like methods by brokers) show the state policing recruitment abuses rather than ordering them; the Yoshida fabrication and the 2014 Asahi retraction; the 2014 Japanese government review of the Kono Statement's drafting, which documented Korean textual input and the absence of documentary corroboration behind some of its language; and US military interrogation materials, above all the September 1944 US Office of War Information Report No. 49 on twenty Korean comfort women captured in Burma, which described them as "nothing more than a camp follower or a professional 'camp follower'" attached to the army, earning wages, able in some cases to refuse customers and to leave after debt repayment [TBD-VERIFY: Report 49's exact phrasing; the document is the single most-cited primary source in the revisionist canon and its representativeness is centrally contested].

Core claims. (1) The system was wartime licensed prostitution under military regulation, tragic and degrading — most serious figures in this account concede the women's suffering — but continuous with the licensed-prostitution regimes then legal in Japan and Korea, not a state enslavement programme; the operative analogy is military-regulated brothels of the kind many armies historically used, not the Atlantic slave trade. (2) The numbers are inflated by an order of magnitude: Hata's recalculation from soldier-strength ratios and station documentation yields roughly 20,000, and the 200,000 figure originated partly in a conflation of comfort women with the Teishintai (Volunteer Labour Corps) — wartime factory-labour mobilisation of Korean women and girls — a conflation present in early Korean reporting and never fully corrected in the advocacy literature [TBD-VERIFY: the genealogy of the 200,000 figure is itself contested; the conflation claim is documented for early-1990s Korean media but advocacy sources derive the figure by other routes as well]. (3) Recruitment was overwhelmingly by private brokers, frequently Korean, operating a pre-existing trafficking economy; where abuses occurred the state's documented posture was regulation, and the absence of any forcible-abduction order after three decades of archival searching is evidence of absence, not absence of evidence. (4) Korea has instrumentalised the issue in a "moving goalposts" pattern: the 1965 treaty settled all claims "completely and finally" and Korea raised no comfort-women claim then; the 1995 Asian Women's Fund offered prime-ministerial apology and money and the movement organised its rejection; the 2015 agreement was "final and irreversible" and was unravelled within three years — from which this account infers that no Japanese concession can ever be final because the issue's function in Korean politics is to remain open, serving domestic nationalism and, in the harder version of the claim, the political economy of the advocacy organisations, a version that received its exhibit in the 2020 Yoon Mee-hyang scandal. (5) The Asahi retraction vindicates scepticism: the most graphic perpetrator testimony in the canon was fabricated, was reported uncritically for three decades by Japan's leading liberal newspaper, contaminated the UN documentation, and shaped the international image of Japan — demonstrating, on this account, that the dominant narrative was constructed with reckless disregard for verification and should be re-examined root and branch.

The account's internal registers. The corpus records that this account is not monolithic. Hata accepts the Semarang case as genuine coercion and condemns it; he accepts military involvement in regulation and logistics; his dispute is over characterisation and scale. The Yomiuri's position has been that the issue was real but the "sex slave" framing and the numbers are wrong. The harder polemical wing — which extends to denial of meaningful suffering, harassment of scholars (the litigation and threats around Uemura Takashi, the former Asahi reporter, and the 2014–2015 attacks on US textbook treatments), and the "history wars" lobbying against overseas statues — is part of the account's political reality and is documented here without being treated as its strongest form.

Weakest points. (1) The Yoshimi documents stand: military establishment, regulation, transport, and medical administration of the stations are documented beyond dispute, and the official pre-1992 position that the system was purely private was false — a falsity exposed not by foreign pressure but by Japanese archival scholarship, which complicates the account's claim that the narrative is a Korean construction. (2) The Semarang convictions establish judicially that forcible enslavement by Japanese officers occurred where the victims were European and the records survived; the inference that comparable episodes occurred against Asian women whose cases produced no post-war tribunal — in China's war zones especially, where survivor testimony of violent seizure is extensive — is at minimum strong, and the account's heavy reliance on the absence of Korean-theatre abduction orders sits awkwardly beside the documented destruction of records in August 1945. (3) The narrow-coercion test proves too little: even granting that most recruitment ran through brokers and deception, a system in which the military created demand for women it then confined in regulated facilities they could not freely leave satisfies most definitions of trafficking and forced prostitution in both contemporary and period international law (Japan had ratified the 1921 trafficking convention, albeit with a colonial-territories reservation [TBD-VERIFY: scope of Japan's reservation]); the moral and legal weight survivor testimony carries does not depend on the abduction question. (4) The account's numerical low end has its own extrapolation problems, and its flagship primary source (OWI Report 49) describes one captured group in one theatre in 1944 under interrogation conditions, a thin base for system-wide characterisation. (5) Politically, the account's victories have been pyrrhic: the 2007 cabinet statement produced House Resolution 121; the post-2014 "history wars" campaigns abroad consolidated rather than dissolved the international consensus around the survivor-centred framing, as reflected in the museum, memorial, and curricular landscape outside Japan.

5. Account 3: The Diplomatic-Realist / Reconciliation-Studies Account

Evidentiary foundations. The third account is grounded not in the wartime record but in the post-1991 diplomatic record: the negotiation histories of 1993, 1995, and 2015 (including the December 2017 Moon task-force report, the most detailed official reconstruction of the 2015 talks); the documented US brokering role (the Obama administration's sustained pressure across 2013–2015, the March 2014 Hague trilateral, and the explicit US framing of Japan-Korea friction as a strategic liability against North Korea and China); and the academic literature of apology and reconciliation politics, above all Jennifer Lind's Sorry States: Apologies in International Politics (2008) — which argued from the Japan-Korea and Franco-German cases that apologies help only when they avoid triggering domestic backlash in the apologising state, and that backlash-provoking contrition can leave bilateral relations worse than restraint — and Kimura Kan's The Burden of the Past: Problems of Historical Perception in Japan-Korea Relations (2014; English 2019), which traced how the issue's salience is generated by the two countries' domestic political cycles rather than by new historical facts. Cheol Hee Park's and Brad Glosserman/Scott Snyder's work (The Japan-South Korea Identity Clash, 2015) supply the identity-politics variant; Onuma Yasuaki's insider account of the Asian Women's Fund supplies the practitioner's post-mortem.

Core claims. (1) The issue is caught in a structural trap with two interlocking mechanisms: on the Japanese side, an apology-fatigue cycle in which each act of contrition (1993, 1995, 2015) provokes a domestic conservative backlash that produces qualifying statements, textbook pressure, or Yasukuni-adjacent gestures, which Korea reads as proof of insincerity, devaluing the apology retroactively; on the Korean side, a ratification-failure cycle in which no Korean government has been able to deliver domestic acceptance of any settlement — 1965 (negotiated by an authoritarian government and repudiated in democratic memory), 1995 (rejected under movement pressure), 2015 (reviewed and gutted within two years) — because the issue is interwoven with the legitimacy politics of the post-colonial Korean state, in which being seen to close accounts with Japan cheaply is politically lethal. Neither mechanism requires bad faith; both run on ordinary democratic politics. (2) Both governments are repeatedly undone by their own domestic politics, not by each other: Kono's statement was followed by LDP-right mobilisation against it; Murayama's Fund was sabotaged jointly by Japanese conservatives (who resented it) and Korean activists (who rejected it); Park concluded the 2015 deal while concealing parts of it from her own public, months before her collapse; Abe paired the agreement with domestic signalling ("we did not admit legal responsibility"; his January 2016 Diet statements re-stating the no-forcible-abduction-documents position) that stripped it of reconciliation value in Korean eyes. (3) The 2015 agreement was diplomatically sound but politically unanchored: judged as text, it was the strongest package ever assembled — prime-ministerial apology, acknowledgement of military involvement and government responsibility, state budget funds (not private donations, the precise defect of 1995), and mutual finality; judged as process, it violated every precept of settlement durability — no survivor participation, no public preparation, unwritten side-understandings bound to leak, and finality language that converted future friction into breach accusations. (4) Memory is managed, not resolved: the realist account holds that deep reconciliation on this issue is unavailable on any policy-relevant horizon, and the achievable objective is insulation — keeping the memory dispute from contaminating security and economic cooperation — which is what the 2015 deal, the Yoon-era quieting, and the Camp David architecture each attempted by different means. (5) Geopolitics drives the cycle's amplitude: the issue flared when the strategic environment permitted (the 2012–2015 trough coinciding with Korean China-engagement under Park) and quieted when threat perceptions converged (post-2022 North Korean missile tempo, the China question, and US pressure producing the 2023 reset); the survivor numbers' decline toward zero will remove the issue's living anchor while, on this account's prediction, leaving the symbolic conflict fully available for future mobilisation.

Weakest points. (1) The account's symmetrical structuralism can shade into false equivalence: it treats a state that operated the wartime system and a state whose citizens were its victims as symmetric players in a memory game, a framing both other accounts reject — the survivor-centred account because it erases the justice asymmetry, the revisionist account because it concedes the underlying characterisation. Structural analysis explains the cycle; it does not adjudicate the claim, and its practitioners sometimes write as if explaining the cycle were adjudicating the claim. (2) Its core prescription — manage, insulate, lower expectations — is normatively thin where the issue is most acute: it has no account of what is owed to the women themselves, and its implicit counsel in 2015 ("take the deal; nothing better will come") asked the survivors to absorb the cost of alliance management, which is precisely the procedural wrong the first account identifies. (3) Lind's backlash thesis is contested empirically: the Kono Statement and the AWF arguably failed not because apology is structurally futile but because each was under-executed and immediately undercut — a quality-of-implementation explanation, not a structural one, and the German comparison (sustained, institutionalised, education-embedded contrition) suggests the cycle is escapable with sufficient political investment, which Japan's leadership has not chosen to make. (4) The account under-predicted the Yoon-era reset's depth and the Moon-era unravelling's limits in both directions [TBD-VERIFY: this historiographical scoring is the corpus's own judgement and should be checked against the account's published predictions], suggesting its cycle model is descriptive rather than predictive. (5) Its reliance on official negotiation records gives it a state's-eye view in which civil society appears mainly as a spoiler — an analytical choice that systematically discounts the possibility that the survivor movement was not noise in the system but the system's conscience, and the reason any acknowledgement exists at all.

6. The 2015 Agreement as Test Object: Three Readings of One Deal

The 28 December 2015 agreement is the document's test object because it is the single artefact about which the factual record is least disputed — the texts of both foreign ministers' announcements are public, the implementation record is documented, the unravelling is datable — so that the three accounts' divergent readings of it isolate their interpretive machinery from factual disagreement.

What each account sees in the same text. The survivor-centred account reads the announcement's omissions: no admission of legal responsibility (the text says "painfully aware of responsibilities", a formula continuous with 1995); no Diet act or cabinet decision embodying the apology; compensation routed through a foundation rather than paid as of right to claimants; the silencing clause; and the survivors' absence from the room — and concludes the deal was an alliance-maintenance transaction denominated in the survivors' claim. The revisionist account reads the same text as Japan's third payment on a bill twice settled: ¥1 billion of state funds and a prime-ministerial apology for a system whose characterisation Japan still contests, extracted under US pressure, in exchange for finality that Korea abandoned within twenty-four months — confirming that concession purchases nothing and that the 2007 position (sympathy without legal concession) was the sustainable one. The realist account reads the deal's architecture: the deliberate symmetry of parallel statements (avoiding a single signed instrument that would require Diet or National Assembly scrutiny), the constructive ambiguity on the statue, the finality clause as mutual political cover — and judges it skilled diplomacy whose failure mode was foreseeable and procedural: settlements that survive are socially pre-ratified, and this one was concluded in secrecy by two leaders who each then declined to spend capital defending it, one of whom was politically destroyed within a year for unrelated reasons.

What each account's reading predicts, and what happened. The survivor-centred reading predicted rejection, and rejection occurred — but the account must also absorb the fact that a substantial majority of the survivors alive at the agreement date accepted the foundation's payments [TBD-VERIFY: mid-thirties of forty-seven], a datum suggesting the movement's rejection and the survivors' individual choices were not identical, and one that complicates the claim that the deal failed because survivors rejected it rather than because the movement and then the Moon government did. The revisionist reading predicted unravelling, and unravelling occurred — but the account must absorb the fact that Korea never returned the ¥1 billion, never formally abrogated the agreement, and that two successive Korean administrations (Yoon, and thus far Lee [TBD-VERIFY]) have treated it as standing, which fits poorly with the pure-instrumentalisation thesis and better with the realist picture of a deal too costly to either honour fully or kill formally. The realist reading predicted that the deal would stabilise the relationship's floor even while failing socially — and the 2023–2025 reset, conducted on top of the formally surviving agreement without renegotiating it, is the strongest evidence in the realist column: the agreement failed as reconciliation and partially succeeded as insulation.

The unwritten-understandings problem. The December 2017 Moon task-force report's disclosure that the negotiation included unannounced understandings — Korean acknowledgement of Japanese concerns on the embassy statue and on third-country memorials, and on avoiding the term "sexual slavery" in official usage — is read three ways: as proof of bad-faith secrecy that vitiated consent (Account 1); as proof that Korea undertook commitments it never intended to keep (Account 2); and as ordinary diplomatic practice whose exposure, not whose existence, was the error — constructive ambiguity being load-bearing precisely until it is published (Account 3). The corpus notes that all three readings are textually available from the same task-force report, which is itself a contested document: Japanese officials disputed aspects of its characterisation [TBD-VERIFY: specific Japanese government responses to the task-force report's findings].

The finality clause as the deal's fossil record. "Final and irreversible" (choejong-jeok mit bulgayeok-jeok / saishū-teki katsu fukagyaku-teki) was reportedly a Korean-side drafting contribution originally aimed at making the Japanese apology irreversible — i.e., at preventing the post-apology backlash cycle — before being turned around to make the settlement irreversible [TBD-VERIFY: this drafting genealogy is reported in Korean accounts of the negotiation and in the task-force report's orbit; the precise authorship is not established in public records]. If verified, the clause's career is the conflict in miniature: each side attempting to bind the other's future politics, neither able to bind its own.

7. The Three Accounts in Dialogue: Convergences, Divergences, and the State of Play, 2026

Convergences. The accounts agree on more than their partisans advertise. All three accept that comfort stations existed across the empire with military involvement in their establishment, regulation, and logistics (the post-1992 documentary floor); all three accept that the Semarang case involved adjudicated forcible enslavement; all three accept that women in the system suffered gravely, that the Yoshida testimony was fabricated, that the 200,000 figure is an estimate rather than a count, and that the 1991–2015 sequence of Japanese responses (Kono, AWF, 2015) each exceeded its predecessor in explicitness and state involvement. All three accept — though they weight it differently — that the issue's career has been shaped at every turn by forces external to the historical question: Korean democratisation, Japanese conservative identity politics, the US alliance system, and the China factor.

The load-bearing divergences. Three disagreements do almost all the work. First, the definition of coercion: whether the system's character is fixed by the recruitment moment (the revisionist test) or by the condition of servitude (the survivor-centred test) — a definitional choice that is logically prior to evidence and that no archival discovery can settle, which is why thirty-five years of research have narrowed the factual range without narrowing the characterisation dispute. Second, the theory of responsibility: whether the 1965 treaty's "completely and finally" language exhausted state-to-state and individual claims (the consistent Japanese legal position, applied to comfort women from 1991 onward), or whether grave human-dignity violations generate non-extinguishable individual claims (the Korean Constitutional Court's 2011 premise and the January 2021 Seoul judgment's holding) — a dispute now lodged in the international law of sovereign immunity, where the 2012 ICJ judgment in Jurisdictional Immunities (Germany v. Italy) supports Japan's position and the Italian Constitutional Court's 2014 response and the Seoul judgments mark the counter-current [TBD-VERIFY: status of any enforcement measures in the Korean proceedings]. Third, the theory of settlement: whether a justice claim can be closed by inter-state agreement without claimant consent — the question on which 1965, 1995, and 2015 each foundered, and on which the three accounts give, respectively, the answers no, yes-and-it-was, and only-if-socially-ratified.

The historiographical state of play. As of 2026 the scholarly centre of gravity outside Japan — in English-language Japan studies, Korean studies, and the international-law literature — works broadly within the survivor-centred account's factual framework (military involvement, multi-form coercion, "sexual slavery" as at least a defensible characterisation) while increasingly absorbing the realist account's analysis of the diplomacy; Yoshimi's corpus remains the documentary baseline and Hata's the principal empiricist counter-reading that serious work must engage. Within Japan, the government position (Kono Statement maintained; "sexual slavery" rejected as inaccurate; 2015 agreement in force; all claims legally settled) has been stable across LDP cabinets since 2015 [TBD-VERIFY: any modification under post-Ishiba cabinets]. The 2021 International Review of Law and Economics article by J. Mark Ramseyer ("Contracting for Sex in the Pacific War"), which modelled the system as voluntary indentured contracting, produced the decade's largest scholarly controversy — mass rebuttals documenting source problems, retraction demands that the journal declined, and a literature of response (Amy Stanley, Hannah Shepherd, Sayaka Chatani, David Ambaras, Chelsea Szendi Schieder's joint critique among them) — which functioned as a stress test the revisionist account's strongest academic version did not pass in the international scholarly forum, while simultaneously demonstrating that account's continuing institutional vitality [TBD-VERIFY: final disposition of the Ramseyer article and journal statements]. In Korea, Park Yu-ha's Comfort Women of the Empire (2013) — which complicated the binary by documenting broker agency, survivor heterogeneity, and Korean collaboration, and was met with criminal defamation prosecution — ended with the Korean Supreme Court's October 2023 acquittal [TBD-VERIFY: final disposition], an episode each account also reads its own way: as revisionism by other means, as vindication of inquiry against orthodoxy, or as evidence that Korean memory politics polices its own dissenters as Japan's polices its own.

The demographic horizon. The registered Korean survivors number in the single digits [TBD-VERIFY: count as of mid-2026]; the Chinese, Filipina, Indonesian, and Taiwanese cohorts are similarly near exhaustion [TBD-VERIFY]. All three accounts agree the issue will outlive the survivors — and disagree about what that means: the survivor-centred account is institutionalising memory against the silence's return (museums, archives, the contested UNESCO Memory of the World nomination [TBD-VERIFY: status of the comfort-women-records nomination, blocked amid Japanese objections from 2017]); the revisionist account expects the issue to lose evidentiary anchor and political force; the realist account expects it to persist as a fully symbolic resource, cheaper to mobilise and harder to settle precisely because no living claimant's consent will be available to confer or withhold legitimacy on any future arrangement.

8. Conclusion

This document has presented the comfort-women issue and the 2015 agreement through three accounts — survivor-centred/transnational-justice, Japanese conservative/revisionist, and diplomatic-realist/reconciliation-studies — and has declared no winner, because the corpus's judgement is that the accounts diverge primarily on definitional and normative commitments (what counts as coercion; what discharges responsibility; what makes a settlement final) that evidence constrains but cannot resolve. The factual spine is no longer seriously disputed in its core: the stations existed, the military built and ran the framework, women from across the empire were held in them under conditions ranging from exploited prostitution to adjudicated enslavement, and the numbers are unknowable within a wide band. What remains disputed is everything the facts are for.

Three closing observations bound the analysis. First, the issue's career falsifies the assumption — shared in 1965, 1995, and 2015 — that state-level settlement closes history; each closure manufactured the grievance that reopened it, and any fourth attempt will face the additional problem that the claim-holders will no longer be alive to accept it. Second, the 2015 agreement remains, formally, in force: neither abrogated nor implemented, criticised by both publics and relied upon by both governments — a condition the realist account would call equilibrium and the other two would call unfinished. Third, for the wider Japan corpus, this file is the paradigm case of Block J's method: where JP-J-03 documents how a shrine became a sensor of historical-memory politics, this document records how a population of women became, successively, invisible, testimonial, contractual, and demographic — and how three communities of interpretation, each internally coherent and each with documented weaknesses, continue to read the same record toward irreconcilable conclusions. The corpus carries all three forward, with their TBD-VERIFY tags open, into the post-survivor era the issue is now entering.

Sources

Yoshimi Yoshiaki, Comfort Women: Sexual Slavery in the Japanese Military During World War II, trans. Suzanne O'Brien (Columbia University Press, 2000; Japanese original Jūgun Ianfu, Iwanami Shoten, 1995).
Yoshimi Yoshiaki (ed.), Jūgun ianfu shiryōshū [Documents on the Military Comfort Women] (Ōtsuki Shoten, 1992).
Hata Ikuhiko, Comfort Women and Sex in the Battle Zone, trans. Jason Michael Morgan (Hamilton Books, 2018; Japanese original Ianfu to senjō no sei, Shinchōsha, 1999).
Statement by Chief Cabinet Secretary Yohei Kono on the result of the study on the issue of "comfort women", 4 August 1993 (MOFA official translation).
Announcement by Foreign Ministers of Japan and the Republic of Korea at the Joint Press Occasion, Seoul, 28 December 2015 (MOFA and ROK MOFA texts).
Radhika Coomaraswamy, Report of the Special Rapporteur on Violence against Women, Addendum: Report on the mission to the DPRK, ROK and Japan on the issue of military sexual slavery in wartime, UN Doc. E/CN.4/1996/53/Add.1 (January 1996).
Gay J. McDougall, Special Rapporteur, Contemporary Forms of Slavery: Systematic Rape, Sexual Slavery and Slavery-like Practices during Armed Conflict, Appendix (UN Doc. E/CN.4/Sub.2/1998/13, June 1998).
Republic of Korea Ministry of Foreign Affairs, Task Force Report on the Review of the Korea-Japan Agreement of December 28, 2015 on the Issue of "Comfort Women" Victims (27 December 2017).
Government of Japan, "Details of Exchanges between Japan and the Republic of Korea (ROK) Regarding the Comfort Women Issue — From the Drafting of the Kono Statement to the Asian Women's Fund" (review report, 20 June 2014).
Asian Women's Fund, Digital Museum: The Comfort Women Issue and the Asian Women's Fund (final reporting and document archive, awf.or.jp).
Asahi Shimbun, "Thinking about the comfort women issue" special review and retraction notices, 5–6 August 2014; Third-Party Committee report on Asahi's comfort-women coverage (December 2014).
Onuma Yasuaki, Ianfu mondai to wa nan datta no ka [What Was the Comfort Women Issue?] (Chūkō Shinsho, 2007).
Jennifer Lind, Sorry States: Apologies in International Politics (Cornell University Press, 2008).
Kimura Kan, The Burden of the Past: Problems of Historical Perception in Japan-Korea Relations, trans. Marie Speed (University of Michigan Press, 2019; Japanese original 2014).
Brad Glosserman and Scott A. Snyder, The Japan-South Korea Identity Clash: East Asian Security and the United States (Columbia University Press, 2015).
C. Sarah Soh, The Comfort Women: Sexual Violence and Postcolonial Memory in Korea and Japan (University of Chicago Press, 2008).
Park Yu-ha, Comfort Women of the Empire (Teikoku no ianfu; Korean original 2013, Japanese edition Asahi Shimbun Shuppan 2014).
Constitutional Court of Korea, Decision 2006Hun-Ma788, 30 August 2011 (failure-to-act on comfort-women claims unconstitutional).
Seoul Central District Court, Judgment of 8 January 2021 (Case 2016Ga-Hap505092, twelve plaintiffs v. Japan); Seoul Central District Court, Judgment of 21 April 2021 (parallel suit dismissed); Seoul High Court appellate judgment, 23 November 2023 [TBD-VERIFY: citation details].
US House of Representatives, H.Res. 121 (110th Congress), agreed 30 July 2007.
Netherlands Government, Report of a Study of Dutch Government Documents on the Forced Prostitution of Dutch Women in the Dutch East Indies during the Japanese Occupation (presented to Parliament, 1994).
US Office of War Information, Psychological Warfare Team, Japanese Prisoner of War Interrogation Report No. 49 (Burma, September–October 1944).
Women's International War Crimes Tribunal on Japan's Military Sexual Slavery, Judgment (The Hague final judgment, 4 December 2001).
Amy Stanley, Hannah Shepherd, Sayaka Chatani, David Ambaras, and Chelsea Szendi Schieder, "'Contracting for Sex in the Pacific War': The Case for Retraction on Grounds of Academic Misconduct" (The Asia-Pacific Journal: Japan Focus, 2021), responding to J. Mark Ramseyer, "Contracting for Sex in the Pacific War", International Review of Law and Economics 65 (2021).

JP-J-03: Yasukuni Shrine and Historical Memory (1978–2025) — the companion contested-legacy file; Yasukuni and the comfort-women issue are the two principal sensors of Japanese historical-memory politics and share the apology-backlash cycle analysed here.
JP-F-03: Japan-Korea Relations from Yoon to Lee Jae-myung (2022–2025) — the bilateral-trajectory file into which this document's post-2022 quieting feeds; the forced-labour third-party-reimbursement formula is that file's centrepiece and this file's nearest structural analogue.
JP-H-PM-02: Shinzo Abe — Biography — the 2007 cabinet statement, the 2014 Kono-Statement review, the 2015 agreement decision, and the domestic management of the apology are read there through Abe's career.
JP-K-04: The 2014–2015 Collective Self-Defense Reinterpretation and Security Legislation Decision — the contemporaneous security-policy decision whose US-alliance logic (Washington's pressure for Japan-Korea functionality) directly conditioned the 2015 agreement's timing.
JP-A-01: The Yoshida Doctrine and the 1955 System — the post-war settlement architecture, including the 1965 normalisation's place in the US-led regional order, within which the claims-settled position originates.
JP-A-02: Abe's First Premiership (2006–2007) — the March 2007 cabinet statement and House Resolution 121 episode in its domestic political context.
JP-B-03: back-reference added by symmetry sweep
JP-C-03: back-reference added by symmetry sweep
JP-C-04: back-reference added by symmetry sweep
JP-F-01: back-reference added by symmetry sweep
JP-N-01: back-reference added by symmetry sweep
JP-R-01: back-reference added by symmetry sweep
JP-A-00: back-reference added by symmetry sweep