Earlier this month at the request of the editor of the International Review of Law and Economics, we began to write a critical response to the article “Contracting for Sex in the Pacific War,” by Professor J. Mark Ramseyer, at that point released online by the journal with plans for formal publication in March.

As historians of Japan and Korea, what initially appalled us was Ramseyer’s elision of the larger political and economic contexts of colonialism and gender in which the comfort women system was conceived and implemented, and the multiple and brutal ways in which it affected and afflicted the women on a human scale. But as we began to look into the article, its evidence, and its logic, we encountered a different and prior problem of the article’s scholarly integrity. We write to explain that problem.

Ramseyer’s article rests on a comparison of contracts concluded with the so-called “comfort women”, mainly Korean women, between 1938 and 1945, with contracts for what we might call ordinary legalized prostitution in prewar Japan and in colonial Korea. The article states that he bases his comparison on examination of all these categories of contract (p. 2, final paragraph of section 1). Yet, so far as we and other scholars can determine from tracking Ramseyer’s citations, he has not consulted a single actual contract concluded between a Korean comfort woman, or her family, and a recruiter or a comfort station, or even a sample contract that might have been provided for guidance by the Japanese government or military. One of his sources (Naimusho 1938) provides sample contracts for Japanese women recruited to comfort stations in Shanghai. It describes the women as shakufu (barmaid) not ianfu (comfort woman). It is written in Japanese.

Absent evidence of contracts concluded in Korea with Korean women, readers are being asked, with no justification given, to assume that such contracts were the same as these contracts with these Japanese women. We do not see how Ramseyer can make credible claims, in extremely emphatic wording, about contracts he has not read.

In addition to the absence of contracts, he offers virtually no documented third-party statements, oral or written, about contracts with Korean women. The final sentence of section 3.2 (p.6) claims that “some Korean comfort women in Burma worked on contracts as short as six months to a year.” The citation brings one to a sample contract written in Japanese in 1937 (years before the Japanese military was fighting in Burma). It is a sample for contracting with Japanese prostitutes to work in “comfort stations” which specifies a two-year term.

There is only one verifiable reference in the entire article to a third-party claim about contracts with women from Korea (section 3.4, final paragraph). Ramseyer refers to a diary kept by a “Korean receptionist for comfort stations in Burma and Singapore,” said to make clear that “regularly, comfort women from his brothel completed their terms and returned to their homes.” He cites a book about that diary, not the diary itself (the diary was translated into Japanese in 2013).¹ In the diary one finds seven entries noting cases where one or two women completed their terms. Most of them applied for permits to return home, but whether all succeeded is not clear. One also finds an entry noting that two women who had left the station by marrying (one assumes, to Japanese officers) were forced to return to their “comfort stations” by a military official.

The same paragraph that mentions the diary, purportedly a paragraph about contracts with Korean women, ends by quoting a Japanese veteran who recruited women from Japan and claimed many of them paid off their advance and went free.

Any reasonable standard of academic integrity would require that Ramseyer state in his article that he does not have access to actual contracts or sample contracts concluded with Korean women in Korea, acknowledge how few third-party statements he has seen about contracts, and note the limits to what one can learn from those references.

For us, as we believe for the journal and for Ramseyer, the heart of his narrowcast argument about contracts rests on the comparison between two types: those concluded with Korean “comfort women” recruited to wartime “comfort stations,” and those concluded with Japanese or Korean women working as prostitutes in prewar licensed brothels in the home islands or the colony. Just as he is unable to make the comparison in the first place, we are unable to critique that comparison with full confidence without having contracts to examine.

Let us explain why seeing the Korean contracts in full text matters so much, beyond the obvious fact that responsible scholarship requires one to be clear on what one’s sources are or are not (we have little doubt such contracts were concluded; the issues are whether samples or concluded contracts survive in any form, and if so, whether Ramseyer’s article points us to any of them).

The word used from 1938 for “comfort stations” (the places the women were put to work) was wianso in Korean, ianjo in Japanese (the same Chinese characters are used in both cases: 慰安所). The term for “comfort woman,” in use from that year, has two of the same syllables/characters, translated as “comfort”: wianbu in Korean, ianfu in Japanese; 慰安婦 in Chinese characters.

So far as we can determine, “comfort woman” (wianbu/ianfu) is a wartime neologism, and “comfort station” is a repurposing of a term that until the late 1930s carried very different meanings. The Asahi, one of Japan’s leading papers, used the term in 9 articles between 1917 and 1935, most with the meaning of “recreation area,” such as a 1930 story celebrating 15 new “comfort stations” (ianjo) in Tokyo parks for the enjoyment of all residents.² The headline of an article in praise of a Japanese hotelier in Seoul who has replaced his shabbier inn with a fine new hotel, published in a Japanese newspaper based in Korea in 1937, calls it “a great advance for ianjo in the [Korean] peninsula.”³ A review of Korean-language newspapers between the 1920s and 1945 shows that the term wianso also held different meanings (e.g., shelters for children, inns and hotels, hot springs spas) in colonial Korea as well, and the term wianbu (慰安婦) begins to appear only in the late 1930s.⁴ A Korean doctoral dissertation from Sŏnggyungwan University in Seoul on the comfort women system (2010) states that “most Koreans did not know what the term wianbu meant.”⁵ And, even a former Japanese military policeman assigned to guard duty at a “comfort station” in 1943 has said that until he got there, he thought he was assigned to an officer’s club, not a brothel.⁶

It matters greatly that the terms now in widespread use in Korean and Japanese to refer to brothels and the women put to work there did not necessarily carry the meanings of brothel or prostitute at the time the Japanese government authorized and arranged for the creation of “comfort stations” and issued instructions to recruit “comfort women.” It means that in oral communication to the women and their families, it was an easy matter to obscure the nature of the work being asked for. Indeed, one finds much oral testimony from the women that they were deceived as to the nature of their expected work. It would be all the more significant if, as we suspect, the contracts themselves used these opaque terms. Of course, we cannot be sure if they did, if neither sample nor actual contracts survive.

The obfuscation of this issue created by the lack of any discussion of whether he has seen actual or sample contracts, and the lack of any citation to such contracts, is for us the most egregious violation of academic integrity in the article. But there are numerous other serious problems: citations that are wholly unrelated to claims made in the text (just one is noted above); claims in the text of the article entirely at odds with the documents cited to support those claims; selective use of documents and other materials to the exclusion of evidence to the contrary. Some of our historian colleagues, including those far more knowledgeable than we on these issues, are compiling an extensive list of such problems. They will be shared with the journal in due course, or may have been shared by the time of this statement, and we believe our colleagues will make that list public.

It is not our responsibility to conduct a full examination of the integrity of a paper published by a journal with which we have no connection. That is the job of the journal and its publisher, ordinarily through the peer review process but in extraordinary cases after the fact. This is such a case. We have written to the journal requesting they suspend publication of this piece, conduct its own inquiry drawing on expert opinion, and pending the result, retract the article.