Abstract

When the US Army footage documenting the Korean victims of wartime sex slavery in the Imperial Japanese Army in China was discovered and released to the public in the summer of 2017, some South Korean academics questioned whether it should be understood as new evidence since it was believed that some of the subjects of the moving image had already been seen and located in photographs discovered previously in the 1990s. This reaction seemed diametrically opposed to that of the mainstream media (and social media), which excitedly praised the discovery of the “brand‐new” evidential document depicting those moving women.

This article examines the politics of using film footage excavated from archives long after the original production of the film. Both types of reactions shown toward the “comfort women” footage shared the same attitude, which viewed the footage as a repository containing certain photographic‐mechanical evidence to prove someone's existence in a scientific manner. Such a belief in the scientific nature of camera images bears a striking similarity with positivist historical approaches. Contrastingly, what Kracauer saw in film as a medium was rather the presence of things that are difficult to freeze and solidify. His concept of the “flow of life” describes an affinity with life (in the form of everyday life) that he theorized films possess but photographs do not. This article seeks a visual sociological method to explore film as a medium that embodies the reality of the subalterns and, if circumstances allow, speaks on their behalf.

Introduction

Following the summer 2017 discovery and public release of US Army Signal Corps footage documenting Korean victims of wartime sex slavery under the Imperial Japanese Army in China, some South Korean academics questioned whether it should be understood as new evidence, since some of the subjects in the moving images had already been identified in photographs discovered in the 1990s. This view of the footage seemed diametrically opposed to the agitated reports circulating in the mainstream media and social media praising the discovery of new, documentary evidence depicting those moving women. This article originally grew out of my personal interest in untangling what had driven me to join the project's research team1 as a researcher in charge of locating those moving images: Why did I feel obligated to provide this specific form of records documenting the victims? Was it a means of proving something that the world did not have enough evidence to conclusively establish? Additionally, in response to those other academics’ skepticism: What can a piece of film footage bring to our time that a photograph cannot? Since the showcase, however, the afterlife of the film footage, which had originally been made for US Army wartime intelligence, has become self-regenerating and extended to problematic issues, including the endemic question of how to represent the victims. Along this line of development, my own questions have evolved into a more complex one: If we assume the film, as qualitative data, can speak for a party other than the creator, does it do so on behalf of the victimized subjects in it, or of the researcher who tries to interpret it? If we take the latter to be the case, or at least adopt the view that the former is possible only through the latter, then what would be the degree of truth the film as a document can contain and deliver, and how can a researcher access it?

The film's afterlife, especially if implicated in the real-life documentation of violence and victimization, also poses a question of ethics. In his book on documentary ethics, Bill Nichols (2016: 151) states that “there is no code of conduct, no set of ethical standards that governs all documentary filmmaking,” and this definition justifies his view that the field of documentary filmmaking falls short of the standards of other disciplines such as journalism, sociology, and anthropology. For historians and historical sociologists who deal with film as an archival document, there is a comparable shortcoming, at a similar degree, in the ethical consensus on how to interpret film. Nichols discusses certain conflicts of interest between filmmakers and their subjects, particularly in circumstances where the subjects are supposed to “relinquish any and all rights to how what is recorded of their lives gets used”: while the subjects may claim the right to intervene in how their actual lives are represented, the filmmaker's prerogative to “make the work they envision, with due regard for ethical considerations,” not “the one their subjects expect from them,” also carries significant weight (151). When considering the relationship between the academic interpreter and the subject, these conflicts of interest are resituated similarly in the realm of ethics, especially when the subjects are not physically existent or have less capacity to speak for themselves. What are the limits to the interpreter's discretion to infer the subjects’ actual lives from the way they are represented in the film? If there is probable cause to believe that the subjects’ representation testifies to atrocities in the past, is the excavator-interpreter liable for the (possibly posthumous) exhibition and circulation of the subjects’ filmic representation? Or, conversely, does the excavator-interpreter even have the right to instigate that chain of events?

This article delves for the seemingly ungeneralizable answers to these thorny questions, specifically by tracing back over the very process of archival excavation, traversing recent issues around camera image documents and memory, and theorizing film's capacity to speak. That is, this article seeks to explore film as a medium that embodies the reality of the subaltern and, if circumstances allow, speaks on their behalf in a period distinct from that of the film's original production, even after a long lapse of time.

Collecting Post/Colonial Images from the Archives of a Visualizing Power

Seoul-based Sinologist Paek Wŏn-dam once pointed out that “the heterotaxis of modern Asian historical documents”—by which she meant to suggest that the best place for Asian studies was Asia-related institutes and libraries in the United States—had long deferred the proper representation of Asia's multiple temporalities (IEAS 2009: 91). The location of the “comfort women”2 footage, the US National Archives and Records Administration (hereafter NARA), is indeed testament to the heterotaxis of Asian historical documents. Since the late 1970s, NARA has been the site of hidden archives that yielded secret records about unspoken atrocities, such as the US massacre of Korean civilians during the Korean War, and that have dramatically upended important historical narratives. Over the last two decades, South Korea's government-sponsored institutions and libraries, including the National Institute of Korean History, the National Archives of Korea, and the National Library of Korea, have invested in locating, collecting, and cataloging Korea-related documents in NARA, and as a result, the sight of Korean scholars busily photocopying documents in NARA's reference room in College Park, Maryland, over their summer or winter recess has not been an unfamiliar one until recently, before the COVID-19 outbreak. This disconnect between the loci of past events and their archives—that is, the heterotaxis—is due not only to the fact that the United States was one of the key actors in these same events, but also to the fact that the United States was and is the foremost power for collecting knowledge from and gazing at other parts of the world. NARA, the federal archive that collects and preserves all records and manuscripts created by the US government, can therefore be called an archive that visualizes each and every unseen and invisible bit of knowledge from other parts of the world within the United States’ reach.

However, this archive of visualizing power is not a voracious whirlpool of information but rather a generous benefactor providing information to many. Upon visiting NARA, you are required to verify your identity with a government-issued picture ID, after which they issue you a researcher card enabling you to pull their holdings out from the stacks. If you follow their rules and regulations for dealing with their documents, such as not bringing scanners with automatic feeders and not touching celluloid or photographic documents with your bare fingers, then you are free to photocopy or scan as many unclassified or declassified, public-domain US government documents as you like, at no cost. As a result, not a few researchers from South Korea and Japan complain about their own national libraries and archives’ bureaucracy and secretiveness in providing access, particularly when compared to NARA. The discrepancy between the level of autonomy for Asian researchers accessing NARA's data and the more restricted access to their own national archives’ data is, in a way, due to the respective managing principles of those archives. In the discipline of archival studies, NARA actualizes one of the two major theoretical traditions in modern Western archives—the one espoused by American archivist Theodore R. Schellenberg (1903–1970). Schellenberg, who began his career at NARA in 1935 and later became the program adviser to the Archivist of the United States in 1948, challenged the other major tradition upheld by British archivist Sir Hilary Jenkinson (1892–1961), who emphasized the archivist's “duty to serve the record by continuing the line of [the creator's] unbroken custody” (Stapleton 1983–84: 77). That is, their key difference lay in their respective views of who was most important at each and every stage of archival management, including collection, appraisal, selection, cataloging, preservation, and providing access. While Jenkinson saw the creator's original intention as the most important, Schellenberg prioritized the future “reference and research use” value, or “useability,” of archival collections (78). From Schellenberg's point of view, the stages of appraisal and selection held much more importance than they did under Jenkinson's archival principles (Kim S. 2016: 77), and the archivist had more freedom in evaluating which data should remain in the archives. According to Stapleton, this emphasis on the data's potential for future use sprung from “the American attitude—and [Schellenberg's] own strong feeling—that public records are, indeed, public property” (Stapleton 1983–84: 78). This usability-oriented style of archival management has become NARA's defining principle for managing its holdings.

It is worth noting, however, that the relatively greater freedoms for research at NARA is given to the researcher as an individual. The “Research Room Rules” on NARA's official website state, “Researcher cards can only be issued to individuals if the records they are researching are physically located within the National Archives” (emphasis added).3 At NARA's reference rooms, researchers do not officially represent their nations or institutions but only themselves as individuals. Though a considerable number of foreign government- or institution-affiliated researchers visit every day, each researcher is regarded an individual researcher, undefined by nationality, occupation, or affiliation. This stance on individuality tells us a lot about the structuring logic of the archive itself. As individuals, researchers gradually adjust themselves to the hierarchy that structures archival data, a hierarchy that echoes the US federal government's own structure. According to Schellenberg, at NARA, “record groups were established for records of administrative units of varying status and authority in the government hierarchy” (Stapleton 1983–84: 79). More importantly, researchers as individuals naturalize the hierarchy that structured how data creators and archivists looked into what happened in other parts of the world, where the data were originally created. I call this gradual internalization process that the individual researcher experiences at NARA a facet of liberal governmentality. That is to say, the heterotaxis of Asian documents does not simply mean the collecting power of Western archives. It also signifies the politics of those archives’ geographical locations, which affect the archivist's appraisal, selection, and structuring of archival data. It entails the subsequent naturalization of that geopolitically driven, structural understanding of records of the past and considerably affects the episteme of Asian academia. The “comfort women” documents housed in NARA, therefore, need to be considered in the contexts of archival management, along with that of original creation.

The Positionality of Archives and the Problems of Cataloging

The documents, whether textual or not, collected by Asian scholars from NARA are conveyed to their country of origin and supplement local archives there to remedy the shortage of primary sources in the historiography of modern Asia. In the process, the question whether the documents are the exact records of pertinent historical events occasionally comes under dispute. More often than not, the undeniably excessive reach of US intelligence works to convince people of the documents’ admissibility as information, or possibly as evidence. However, the structural logic of the original archives, parts of which housed those documents in NARA's catalog, is our major concern here. NARA's identification procedures for Korean “comfort women” in US Army film footage show how archival management and the positionality of archives affect the ways excavated documents can or cannot serve the aim of providing relevant information or evidence.

The two pieces of “comfort women” footage released in 2017 were cinematographed by the US Army, and the unit in charge of making that footage was the US Army Signal Corps. Since its establishment just before the outbreak of the American Civil War, the Signal Corps has been overseeing all communication-related activities in the US Army as a whole and has dispatched its soldiers to battle units in the field (Raines 1999: 5–8). The Signal Corps actively made use of machine-powered, visual documentation technologies, including photographs and motion pictures, and during World War II it collected a massive amount of photographic and filmic images from battlefields, former enemy territories, current and former colonies, and occupied areas (Signal1953: 25). While conducting research at the textual and photographic reference rooms in NARA, the research team found that around 1944, the Signal Corps dispatched the 164th Signal Photo Company to the battlefield in the border region between China and Burma. Signal soldiers out in the field usually operated in teams of two, one for photography and the other for cinematography (Pak Ŭ. 2017: 20). Some of the photographic documents and captions collected by the research team indicated that Sergeant Hatfield and Sergeant Fay were among those soldiers who took pictures in the region. Fay was the cinematographer working with Hatfield, who, in turn, left some photos depicting “comfort women.” Working from this information, I started searching for film materials either shot by Fay or shot in the relevant regions, such as Tengchung, Songshan, and Lungling in China; and Myitkyina in Burma. I viewed over seventy reels of Signal Corps–produced films in Record Group 111, “Records of the Office of the Chief Signal Officer,” at NARA, and found two items with several images of what are likely either “comfort women” or “comfort houses.” Lungling Remains, China (RG111-ADC-2738) is an unedited compilation of film footage taken in the border area between China and Burma, Lungling, on November 10, 1944, and presents the image of a so-called comfort house. The footage of this hotel-turned-comfort house was shot by Fay, as evidenced by his name on the clapboard image at the beginning of the footage (see fig. 1). The other footage, depicting images of six female POWs captured at Songshan in 1944, is compiled in World War II in China (RG111-ADC-9706).

The identification of these women and the venue was not a simple task. Since Signal Corps soldiers sent their original negative films to the Signal Corps Photographic Center, later Army Pictorial Center, in Astoria, New York, to process and archive them (Raines 1999: 352), the films’ subsequent archival descriptions seem to have been created usually by collection managers or archivists at the photographic center. Those in charge of creating descriptions were thousands of miles away from the filming locations and had most likely never visited those places. Therefore the descriptions frequently have inaccurate or incorrect information about the subjects in the films. After those films had finished serving as raw materials for secondary production, the films were then sent to NARA to be preserved as data produced by the federal government. The photographic center's descriptions of these films were some of the essential pieces of information that NARA had to collect and apply to their own cataloging. To identify specific figures, locations, or objects in those archived films, therefore, one must consider this context of archiving and these films’ possible trajectories before ending up housed in a stable, governmental archive.

The description in NARA's catalog about the filmed footage compilation World War II in China, which the research team has confirmed as including moving images of living “comfort women,” describes the women as “Chinese girls.” This is a common example of racial generalization in NARA's cataloging, whereby a non-Caucasian-looking subject without relevant, personally identifying information is arbitrarily identified as a member of a specific racial or ethnic group. Keeping such instances of misinformed cataloging in mind, I entered keywords like “Chinese” and “Japanese,” along with “Korean,” into NARA's online catalog, and retrieved a shot list that included the term “Chinese girls” in a description of a medium shot (“MS”) of female subjects in the footage (see table 1). The Signal Corps Photographic Center's description of the “comfort house” film even refers to the women at the venue as “Geisha girls,” freely using the name of Japan's traditional entertainers, who have been objects of fetishism and myth, often misrecognized as prostitutes, especially through Western eyes (H. Kim 2018: 303–4). Therefore, it is highly likely that the women in the moving images were involved in sex slavery or sexual services under the Japanese Army, regardless of their actual ethnicities. We can identify such misinformation by cross-checking with several other documents in different formats. For instance, in the research team's collection of NARA photographs, there is a photo album page with a description using the term “Geisha girls” to designate Korean “comfort women” in Okinawa, also calling them “Jap Korean women” (RG80-G 319133; RG80-G 319155; RG80-G 319156; RG80-G 319157; RG80-G 319163; see fig. 2). Based on textual documents indicating which US Army personnel took charge of photography and cinematography (Kang 2019: 151), the movements of the US-China joint forces in the region (Chŏng C. 2018: 117–288), and the fact that there were Korean “comfort women” in the territory (occupied by the Imperial Japanese Army until the joint forces defeated them) (230–84), the research team and I were able to locate these two pieces of film footage which included images of former “comfort women” believed to be Koreans, along with images of their place of residence.

Photos previously collected by the research team helped presume the ethnicities of several women in the footage. A Signal Corps photo taken from the China-Burma-India (CBI) theater of war, depicting four women who were captured there by the joint forces, has a caption on its reverse side, which seems to have been typed by the photographer at the time of printing, indicating that the women in the photo were “Korean” (RG111-SC 247386). While another photograph taken by the same unit, which was released to the public in the 1990s, has a caption identifying these four women as Japanese (RG111-SC 230147), there was a countertestimony that a pregnant-looking woman in the photo was actually Korean, one who was living in the northern part of Korea at the time. The woman, Pak Yŏng-sim, testified in 2000 that she was the pregnant-looking woman in the picture (Y. Kim 2000). By comparing the iconic proximity of facial features and clothing in these documents of different formats, compounded with supporting testimony from a survivor, the research team concluded that at least two women in the footage compilation, World War II in China, were Korean, with the strong possibility that one other woman was Korean as well.

The process of identifying these female Asian subjects in images held at an American government archive points to issues of racial politics and questions about postcolonial/neo-imperial topography. Who photographed or cinematographed the subjects at these specific moments in history? What was the photographer/cinematographer's racial and/or cultural background, and was it cognate with the photographer/cinematographer's position, angle, and/or capacity to provide additional information about the images? How many, and what kinds of, stages did those photographed/cinematographed images pass through before ending up in NARA, and who was in charge of appraisal and cataloging at each stage? What archival principle informed American archivists’ decisions in selecting, cataloging, and providing access to these images, at least until Asian researcher-excavators eventually reached them? Were the archivists’ cultural backgrounds cognate with the vocabularies, expressions, and/or omissions in the metadata they created? It may be difficult or even impossible to answer these questions for every single document; however, the act of questioning itself and the recognition of NARA's positionality will help the researcher-cataloger at local archives catch any discrepancies between the visual document's actual context of creation and the metadata subsequently created.

The Politics of Optical Evidentiality

When the research team released these two pieces of film footage to the public in July 2017, they immediately garnered immense attention, especially on the internet. Media outlets and anonymous Korean internet users circulated these moving image clips through various online media platforms, accompanied by sensational headlines and, at times, considerably angry comments about the Japanese Army's inhumanity. There were also relatively self-possessed but strategic responses requesting that the government use the footage in diplomatic maneuvers: “There have been many viewers who want this footage to be ‘an ace in the hole’ to get Japan to the table of renegotiation” (Pak U. 2017: 19). At the same time, in some academic groups, there were some markedly different attitudes. A number of academics raised questions about the newly excavated footage's usefulness, so to speak, asking, “What is the use of an additional discovery of film footage, when there is already photographic evidence of the same event?” (19) These academics expressed skepticism about the footage's admissibility as new evidence, largely because they believed that the existing photographic evidence of Pak Yŏng-sim, which was discovered at NARA in the 1990s and then verified by Pak in 2000, was enough to prove the existence of Korean “comfort women” in the Imperial Japanese Army. Alternatively, the sensational ways in which the media used the excavated footage to catch the public's attention might have given these academics a negative impression of the use of moving images.

In fact, these two stances are rooted in the same belief. Specifically, it is the belief that camera image documents, including both photography and film, are scientific evidence of the subject's existence in reality. Many see camera image documents as exact copies of the real world, since they are the products of optical technology, as well as mechanical production and reproduction. In Kim-Gun (2018), a documentary that traces the story of an unknown citizen soldier photographed during the Kwangju Uprising in 1980, this belief in optical evidentiality is called a “quasi-science,” one that has generated ludicrous social controversy over the identity of photographed subjects. In particular, the Far Right political commentator Chi Man-wŏn claims that there was a conspiracy to stoke growing unrest against Chun Doo-hwan's junta and cause a riot by smuggling disguised North Korean soldiers into Kwangju. He and his fellow “professionals” juxtapose reportage photographs of citizen soldiers from 1980 with recent news photos from North Korean media, each of which is matched to an old photo from Kwangju, so as to argue that the photographed citizen soldiers in Kwangju were actually soldiers from the North, who, after the Kwangju Uprising, returned to North Korea and were elevated in status as a reward. The key figure of the documentary, “Kim,” is the first photographed subject that Chi identifies as one of these North Korean soldiers. The documentary crew meets with those who remember Kim, those who were assigned to the same military unit, and those who might themselves be the photographed anonymous soldier, and they try to reconstruct the memory of Kim and his unit in the citizen militia. After putting the pieces of the puzzle together, the crew comes to the conclusion that Kim was a ragpicker who used to live under a bridge in Kwangju and, during the uprising, voluntarily joined the citizen militia. In the end, the documentary confronts the audience with the possibility that Kim was shot to death by Chun Doo-hwan's martial law troops at the end of the Kwangju Uprising.4 Over the course of the investigation, the documentary also portrays other citizen soldiers who realize that Chi has identified them as North Korean soldiers. Having resumed their previous lives as ordinary citizens after the end of the uprising, they find Chi's accusations insulting and bring a libel suit against him. Their attorney, having secured testimonies from the photographer who actually took the photos on May 22 and 23, 1980, expresses her confidence about winning the lawsuit to the documentary crew, saying, “The testimonies have concreteness and coherence that only the person who was at the scene can provide.” To many, camera image documents are considered objective and scientific, as exemplified by the computer display scene, where Chi and his professionals skillfully compare the contour line points of a subject's facial shape with those of his Northern counterpart. However, the veracity of such evidential images can only be supported or refuted by coherent and concrete memories in their full contexts. One of the citizen soldiers who survived the massacre in Kwangju fires back, “Isn't the act of collecting evidence and testimonies to refute the suspicion that we were fake the same act they commit against us in presenting certain evidence?” This question is also relevant to the preference for evidentiality in discussions of the “comfort women” footage's usefulness. Why is it important to provide a “brand-new” set of evidence to prove that “comfort women” really existed in history? Isn't the act of excavating new evidence from NARA to refute Japan's denial of wartime sex slavery the same as Japan's act of denying based on the alleged nonexistence of any written evidence?

The documentary's unrepentant pursuit of truth in spite of the survivor's pessimistic question gives us a clue to the answer to these questions. Kim-Gun begins with an accusation and ends with its dispelling. However, the documentary does not merely focus on how to disprove an accusation's “quasi-scientific” or “fake forensic” (Kim J. 2020: 233) evidence. Rather, it gives more room for the affective contexts of the memories of Kwangju and its aftereffects on survivors’ lives, including the ways that lower-class, orphaned ragpickers won their self-esteem in their fight against state violence, the reasons why so many of them remained anonymous and forgotten even after being victimized, and the ways that the survivors’ agony and trauma linger on. And in the process, those photographs that previously seemed like mere phlegmatic proof of someone's existence gradually disclose the subject's affects and emotions. The subject in question, Kim, and some of his fellow citizen soldiers of similar social backgrounds seem to have collectively identified as being marginalized and unnoticed in society, so their service in the citizen militia might have been an important source of self-esteem and dignity as members of Kwangju society. Kim's sharp and wary eye in the photograph, as the photographer himself speculates in the documentary, might have been an expression of his displeasure at the photographer's unauthorized shot, which not only put Kim in jeopardy by potentially exposing him to government agents but also dethroned the power of the gaze, replacing Kim's reconnoitering gaze as a citizen guard with the camera eye mechanically recording the scene. By providing such fleshed-out descriptions of Kim throughout the whole film, the documentary successfully presents a more comprehensive picture of the photographed scene. That Kim was a real citizen of Kwangju is just a small part of the picture.

Aside from the fact that camera images have been “easily manipulated or altered” to prove the subject's existence or nonexistence (Sturken and Cartwright 2001: 17), we need to question the nature of the “truth” that a camera image can deliver to the viewer. Even in cases where photographs were “actually taken” on the very site of atrocities, to say they are true “tells us very little about the way in which they communicate an understanding of the devastation to those who did not experience it” (Morris-Suzuki 2005: 82). In this sense, the identification of the Korean “comfort women” victims in NARA's film footage contributes very little to our knowledge of what happened to them and how they experienced that time. The evidentiality of the seemingly scientific, mechanically produced Signal Corps film footage was confirmed only by intertextual investigation and, more importantly, with the survivor's own testimony. The iconic proximity of the subjects did not provide the scientific evidence to help this confirmation, because it does not prove the indexical relationship between the image and the subjects in reality. Similarly, Chi Man-wŏn's photographic identification based on the visual comparison of iconic signs was completely refuted by the voices of Kwangju survivors and the rich contexts they provided. In this way, the camera image document's own evidentiality is vulnerable and incomplete without contextual support. If the filmic imagery of those moving, living victims and their place of residence can contribute to a more comprehensive understanding of wartime sexual transgressions, then that contribution must be something more than the fact that they existed there, which I will discuss later in this article.

For that reason, the research team's February 2018 release of another piece of NARA footage, this time depicting dead bodies piled in a pit located among the remains of the Japanese Army in Tengchung, China (RG111-ADC-2417), was, while politically effective and headline-grabbing, contentious in terms of both the act's academic value and the ethics of representation.5 The research team released the death pit footage, along with a US-China joint forces daily operation diary that described how in September 1944, on the “night of the 13th the Japs shot 30 Korean girls in the city.”6 Additionally, the team released the footage at an international conference commemorating the March First Independence Movement, an event cohosted by the Seoul Metropolitan Government on February 27, 2018 (Yi M. 2018). They showcased moving images of dead bodies, taken from NARA footage, before TV reporters and journalists, who were given a press release. The digital copy of the graphic scenes of corpses then began spreading at an alarming rate, first through the YouTube channels of major television news shows and, shortly after, through various social media outlets including Facebook, Instagram, Kakao TV, and Naver TV. Much like the previous NARA footage release, this wide public release seemed to generate the expected effect of reassuring people that there was concrete, visual evidence of “comfort women's” sexual slavery and mass killing committed by the Imperial Japanese Army. At least, this was the consensus among Korean-language users on the web, whereas the Japanese Far Right vehemently denied its admissibility as evidence. After the footage's release, the latter group immediately questioned the footage's capacity to prove that the incident had really happened, especially since the metadata in NARA's catalog (RG111-ADC-2417) provided a shot list identifying the bodies as “dead Japanese soldiers” and “dead civilians, women and children in open pit” (Kim P. 2018). Again, considering the racial politics of NARA's archival management, the identification of the dead bodies as Japanese does not necessarily mean that they were indeed ethnically Japanese, especially in light of the fact that the US Army Signal Corps Photographic Center's archivist very likely misidentified Koreans in the Japanese Army as Japanese in 1944, before the end of WWII. The textual document describing the killing of thirty Korean women,7 together with the visual similarities linking the film footage's presentation of the dead and their clothes with that of other photographic documents collected from NARA (111-SC-212090; 111-SC-212091), both point to the high probability that the footage depicts a real incident against Korean “comfort women” victims. However, aside from the footage's high probability as a record of the incident, one should also consider the effect of releasing such visual documents to the public through the mass media, as these documents do not merely verify the existence of historical figures or events but rather also impact the viewer's affective experience of the visualized scenes.8 While the research team could have easily predicted the controversy over the footage's evidentiality, particularly in light of their own experiences with NARA's cataloging errors and missing information, they also did not seem to treat these documents with any special care, such as, for example, limiting the release to an on-site exhibition or restricting the audiences with access to the graphic images. The widely circulated moving images portrayed the naked bodies of the dead, their genders unidentifiable, piled one on top of another. When we assume that those dead bodies were female, following the inference that they were mostly Korean “comfort women” killed by the Japanese Army, what is the affective impact of the footage's very public release, particularly with regard to the victims’ representation?

First, the showcased and circulated images of “mutilated female bodies” worked to “urge the responsibility of men” and the nation, two parties that had failed to protect women from the cruelty of foreign forces (Y. Kim 2018: 163–64). The use and movement of these images followed a familiar pattern, recalling the 1992 crime scene photographs of Yun Kŭm-i, a victim of fatal abuse by a US Army soldier, and the 2002 graphic images of a US armored car accident that killed two Korean female students. While these images successfully posited men as the primary agents in the struggle against imperialism and/or neo-imperialism, they also rendered the “acute sense of pain the violated female bodies had experienced unspoken or marginalized,” allotting women only a “mediating role” in men's fight (Kim Yŏng-hŭi 2018: 163–64). Such foreclosure of women's agency in framing national narratives reaches its dramatic peak in displaying the victims’ lifeless bodies. In serving as a mediator for men's heroic fight, these spectacular camera images of mutilated female bodies function to bring another effect—“atrocity's normalization” that would generate the viewer's “narcotization” in being “overwhelmed with information” (Zelizer 1998: 212–13). That is, the exhibition of these graphic images has little educational value to prevent a recurrence of such inhumane crimes but rather familiarizes the viewer with the bodily representation of atrocities against women.

In a similar vein, the showcasing of those bodies erases victims’ agency by objectifying their naked, lifeless bodies and limiting the representation of “comfort women” to a strict binary: one as a “familialized” grandmother-activist who fosters solidarity within the bounds of “the same national identity” (Hŏ 2018: 146–51) and the other as an idealized little girl, undefiled and immune to the dangers of sexualization9 and, more importantly, whose “life after victimization” has never been represented (Kwŏn-kim 2019: 68–9). I will discuss this matter later in this article.

Last but not least, the research team's public release again reminds us of the preoccupation with the evidential capacity of camera images, which generates certain political developments disconnected from the affective or representational aspects of showing victims’ images. Since the Japanese government has vehemently denied the Imperial Japanese Army's forced mobilization of Korean women into wartime sex slavery, academic discussions around this issue have long centered on how to prove the crime and what evidence can do so. Pursuing such evidence, one of the research team members claimed that the footage was effectively “the documents that [would] provide information about the circumstances and actual conditions of Korean comfort women in the last stage of WWII, while the Japanese government [was] denying the fact of mass killing” (Yi M. 2018). The same researcher, however, admitted that photographs are not “a faithful depiction of reality” and that there is “always a blind side to the photograph” (Kang 2019: 145). As a way of remedying this blindness, he suggests “a more accurate analysis” by cross-examining with “verified historical facts” (147). This view sums up the research team's basic understanding of camera images as documentary evidence; that is, they see camera images as containers of information, which, because of “the photographer's position,” necessarily leave out or “conceal” some information (145). From this perspective, camera images can serve as evidence only by first being accurately verified. This is a revised version of the belief in optical evidentiality, or in the “truth of an indexical,” to use Allan Sekula's (1986: 6) term, which prioritizes the information value of camera images. This belief entails that camera images from the past are containers of historical information and that the present-day viewer's encounter with them is a matter striving for an “accurate” understanding of the past, specifically by sorting through information from both the container itself and other “verified” sources. Therefore, for proponents of this view, both the ethical issues of representation and the foreclosed agency of women/victims are collateral and insubstantial elements in the process of historicizing the past. The fundamental meaning of camera images is their role as evidence.

Photographer and photographic historian Janina Struk (2004: 15) grimly states that the photographic representation “does not give a comprehensive account of the historical events” and that it is not possible for it to do so because photographic images are mere “fragments” of the past events. What the research team wanted to deliver through the exhibition of the death pit footage was a narrative that could be constructed by absorbing and conflating a series of information, including the facts that the Imperial Japanese Army's sex slavery in Tengchung had existed, that Korean women had been mobilized to serve as sex slaves in the region, and that there had been a mass killing of Korean “comfort women” before the shooting of this footage. However, none of these facts can be proved by the footage itself; other types of documents, such as textual records, photographic records with captions, and the surviving victim's oral testimony, proved them. What the footage delivers is only the visual “fragments” of what happened to exist right in front of the camera at the time of shooting. In other words, the camera image of the death pit cannot function as the evidence of the historical narrative of the atrocity against Korean “comfort women” by the Imperial Japanese Army. What the research team tried to fill the vacancy with was not a “blind side” of the camera image as a container but, according to Struk, a “created memory” or a “fiction” that would invite “fantasy” (213).

Camera images, despite this fundamental limit in giving a comprehensive account, are often welcomed to serve as scientific evidence of the historical events. At a general discussion of how to remember “comfort women,” feminist scholar Chŏng Hŭi-jin criticized this preoccupation with evidentiality, particularly among “mainstream ‘comfort women’ researchers” in South Korea. These researchers have positioned themselves as “standard” experts who collect and provide materials as “objective evidence” in the “division of labor” formed around South Korea's “comfort women” studies, where “standard activists” and “standard victims” coexist with “standard experts” (Chŏng H. 2019). These “standard” roles have formed in both the government-centered, academic financing system and in the nationalization of discourses on victimhood. To help the nation-state in its diplomatic fight against Japan, these “standard” actors have continuously demanded, produced, circulated, and utilized certain types of evidence and evidence-based arguments. Interestingly, their anti-nationalist counterpart also claims to advocate objectivity, as evidenced by the recent controversy over Pak Yu-ha's book Cheguk ŭi wianbu (Comfort Women of the Empire) (Chŏng H. 2019). To counter nationalist narratives of victimization, Pak selectively collected evidence for her hypothesis from testimonies compiled by her antipode, the Korean Council for Women Drafted for Military Sexual Slavery by Japan. This “usurping of testimonies” (Yi H. 2017: 353) has served Pak's “naïve composition of opposition” between “domestic nationalism and cosmopolitan feminism,” which lacks proper consideration of the multifaceted roles of nationalism and feminism in the postcolonial world and has merely “perfected a weightless feminism” (Hŏ 2018: 144).

The emphasis on the evidentiality of camera images, as well as of other types of documents, often limits what the document represents and/or is capable of representing to what is useful for adopting or rejecting certain hypotheses, whether nationalist, anti-nationalist, or even feminist. The questions about the newly discovered footage's admissibility as evidence, the quasi-scientific belief in the photograph's optical evidentiality, and the decision to publicly release graphic images as strong factual data favoring the confirmation of a historical narrative—to some degree, all share the same attitude and ethical stance on the representation of victims in camera images. Once the representation is tried in the court of evidentiality, the most important issue becomes who has the authority to judge a camera image's capacity to deliver evidence proving a past event, and notably not how the representation interacts with the viewer in the present day.

The Camera Image Subject's Capacity to Speak

My question here centers on what a piece of footage can give us beyond the indexical information prioritized by the view of images as containers. Put differently, my question concerns the subject's capacity to speak in camera image documents. I find that there is a recurring correlation between the camera image's evidentiality and its subject; that is, the more a camera image document's nature is emphasized as scientific and measurable, the more the agency of its subject is ignored or erased.

To delve deeper into this, it is worth looking at another case notable for its camera images’ evidential contribution to contemporary South Korean historiography—specifically, the photographs of the 1948 Yŏsu-Sunch’ŏn incident and its aftermath. The incident occurred on October 19, when the 14th Regiment of the South Korean Constabulary (a former branch of the Republic of Korea Army) rebelled in Yŏsu, refusing to transfer to Cheju to suppress a popular uprising on the island. The rebellion was led by leftist officers and soldiers who had opposed the establishment of a separate government in South Korea and afterward had expressed strong dissatisfaction with the newly established government's suppression of the Cheju Uprising. In response, the Syngman Rhee government and the pro-government media steadily produced and circulated “disciplinary knowledge” equating the rebellion with “violence in and of itself,” largely by visualizing the rebel forces’ alleged cruelties and hard-heartedness (Im C. 2005: 116). Among the materials deployed, photographs were frequently used to fill these purposes. South Korea's right-leaning artists’ association, Chŏn'guk munhwa tanch'e ch'ongyŏnmaeng (National Federation of Cultural Organizations), in 1949 published the book Pallan kwa minjok ŭi kago (Rebellion and the Nation's Resolution), which included several photographs and illustrations of the mutilated bodies of civilian victims, who had allegedly been killed by “cruel insurrectionists” (117). Im Chong-myŏng defines the two strategies of visualization used in these photographs as, on one hand, the “numerical representation” that “offered seeming objectivity,” specifically by visualizing the number of victims and losses in the photographs’ metadata, such as their captions, and on the other, the “figurative representation” that “gave shape to gruesome scenes of the incident” and “emphasized reality by appealing to instinctive senses” (112–13). The rebel troops, most of whom had been subdued, executed, purged, or exiled from the army, were saddled with the blame and guilt for the terrible scenes in the Yŏsu-Sunch’ŏn photographs.

Interestingly, a counternarrative to this rightist historicization of the incident has also garnered public attention with the support of photographs. Documentary photographer Yi Kyŏng-mo's collection Kyŏktonggi ŭi hyŏnjang: Yi kyŏng-mo sajinjip (The Scenes of the Turbulent Era: Yi Kyŏng-mo Photo Collection), was first published in 1989 and suggested a new type of historical data for the emerging, progressive historiography that had begun to reexamine the Yŏsu-Sunch’ŏn incident. While Yi viewed the incident from a pro-government journalist's perspective (Kim Tae-sik 1990: 353), his documentary photographs did not fail to document the slaughter of civilians, who the police suspected were rebels or pro-rebel leftists (Yi K. 2010: 76–83). At the time of the book's publication, one reviewer, Kim Tae-sik (1990: 353), attributed a high value to Yi's photo collection, noting, “Whereas a historian's document is ex post facto, a camera's document is immediate on the scene.” Yi's photographs of the victims’ dead bodies and the victims’ family members wailing beside them were received by many, much like Kim described, as the document testifying to the real event that the authoritarian government had concealed. In the same vein, the National Archives of Korea have included his photographs in their official collections, having appraised the photographs as having “archival value” (Chŏng I. 2003). The indexicality of camera images here works to refute the one-sided historicization of the event and to lay the groundwork for a counternarrative. The issue of a camera image's indexical nature, specifically during the Yŏsu-Sunch’ŏn incident, has recently evolved into a legal matter. In March 2019, the Supreme Court approved an application submitted by the bereaved families for a retrial for the death sentence given to three Sunch’ŏn residents who were prosecuted for supporting the rebel forces and, only twenty-two days after their arrest, were executed by the military (Pae M. 2019). The plaintiff's special committee submitted photos taken by then Life photographer Carl Mydans, along with his own memoirs, as key pieces of evidence. A television news report interviewing the committee chief put two photographs by Mydans on air, with the reporter's voiceover narration drawing attention to the dead bodies of civilians left in the bushes, and a couple of soldiers looking at them (Yang 2018).

The mass killings of anti-government army personnel, political prisoners, and civilians before, during, and after the Korean War have been the center of attention in investigations of past atrocities, which have been undertaken as a means of democratizing historiography and overturning unjust convictions in South Korea. As with the Yŏsu-Sunch’ŏn incident, camera images have often provided circumstantial clues, sometimes legal bases, to support plaintiffs’ claims in many investigations. In 1999, a series of photographs discovered in NARA by a New York–based Korean American scholar, Do-young Lee, supported the textual record written by a CIA agent, which claimed that at least two thousand political prisoners in Taejŏn were massacred by South Korean government authorities between July 2 and 6, 1950, immediately after the outbreak of the Korean War (Cumings 2002: 159–60).

However, in most historical narratives based on camera images as evidence, those who were murdered have been represented only as victims or as objectified figures visualizing the violence committed by the assailants. The disclosure of the execution photographs, many from journalists’ photo collections and many others from NARA's excavated collections, have also contributed to the photographed subjects’ victimization. Do-young Lee's excavated photographs offered a powerful visual narrative depicting each step of the execution procedure, including the political prisoners being forced to disembark from a truck, the prisoners lying face down and beaten in the fields, the military executioners pointing their rifles, the prisoners being shot, the executioners checking for survivors, and a right-wing youth group burying the dead bodies in a pit (Pak T. 2017). While this narrative successfully countered the rightist historiography catering to the authoritarian government, it also did not give any answers to the question of the victimized subject's agency.

Here, we need to pay attention not to the idea of these images as containers but rather to the ways they were exhibited and the context in which viewers saw them. While the images were used to support historical hypotheses, the exhibitors saw them exclusively as manifestations of objectivity. The optical machines that captured the scenes were posited as objective witnesses indexically recording the truth and therefore were seen as having transferred their authority as a witness to the excavator-exhibitors in the present day. The rightist historians of the Syngman Rhee regime, who demonized the rebel forces, and the progressive historians of the democratic era (particularly those who proved the victimization of political prisoners and pro-leftist civilians) shared this self-image of upholding objectivity. As such, even though progressive historians sought to free victims from unjust charges, they spoke for the victims only so long as they remained victims.

However, when the viewer tries to see beyond the limits of indexical and evidential information and admits that her/his own subjective position affects her/his interpretation, camera images can occasionally allow for the photographed subject to communicate with the viewer. One example is a 1949 piece of film footage depicting the execution of five Communist officers and soldiers in the Korean Constabulary, taken by an American officer, James H. Hausman. The footage, now in the Hausman Archive at the Harvard-Yenching Library, was shot in Susaek, then a suburb of Seoul, in late 1949 (Kim Tŭk-chung 2001). It opens with a scene of the officers and soldiers being transferred to an undisclosed site by truck. Over the following sequences, the faces of the young officers and soldiers do not just reveal their anxiety over their looming execution but also shift from being serene and unconcerned to animated in laughter over something a Korean officer said—an officer who might have once been their colleague but is now guarding them as enemies (see fig. 3).

In the next scene, where the prisoners get out of the truck and pose for the camera, the man on the left, Lt. Kim Chong-sŏk, who was the leader of an underground Communist organization in the Korean Constabulary, even gives what might be interpreted as a confident smile. Kim shakes hands with an American officer on the right, Captain Hausman, who oversaw the execution, as well as the filming of this footage. In Hausman's own memoir, he mentions that he and Kim were close and that he had trusted Kim until the latter's secret thoughts came to light (Hausman and Chŏng 1995: 191–2). When the two shake hands, Kim is smiling widely (which Hausman seldom does), and he seems to be trying to exhibit self-esteem before his former boss and now executioner (see fig. 4).

This footage gives viewers a glimpse of the already warlike atmosphere in South Korea even in peacetime, just before the outbreak of the Korean War, as well as of how and what those purged in these maneuvers tried to speak in front of the camera. When trying to interpret the men's facial expressions and gestures, it is difficult to limit their identities to that of mere victims. The montage of their scenes, therefore, provides some bits of information that does not necessarily lead to hard evidence or indexical truth. This kind of analysis may seem subjective and less important to those seeking authentic, factual information in camera images; however, as we have seen in the controversy over the ethnicity of dead bodies in the US Army's Tengchung footage, what is presumed factual and indexical still faces the question of its admissibility and still needs to be cross-referenced with other supporting data. Moreover, at this point in the analysis, the crucial element is whether and how the viewer-interpreter's active reading draws out voices from the camera image's subjects, not only through the camera images’ visual data but also through the circumstantial data contextualizing the situation, thoughts and beliefs, and possibly emotions of the subject. Hausman's memoir recalls Kim Chong-sŏk's last words as “both knowable and not knowable” (Hausman and Chŏng 1995: 192), while several historical works on underground activities and the cultures of Communist organizations in colonial Korea document Communist activists’ efforts to maintain their integrity (Yu 1947; Im K. 2002: 269–71). Together, sources like these sketch out the stories of the captured officers and soldiers and provide us with clues for hermeneutic interpretation.

The Flow of Life and Experience: Camera Image as Qualitative Data

The German-born sociologist Siegfried Kracauer once wrote that film has an affinity with life. The positivist framework of science in academia and beyond has been built on the idea that the camera is an “objective” machine that “captures reality” (Sturken and Cartwright 2001: 280) and that a captured moment of time can be scientific evidence. In other words, a camera image is understood as scientific material proving a thing's actual existence at a specific moment in time. Kracauer saw this indexical view as having an affinity with the historical approach championed by Leopold von Ranke, who famously described the historian's task as showing “wie es eigentlich gewesen” (how it really was) (Ranke 1874: vii). What set film apart for Kracauer (1960: 71) was its affinity with life: “One may also say that [films] have an affinity, evidently denied to photography, for the continuum of life or the ‘flow of life,’ which of course is identical with open-ended life. The concept ‘flow of life,’ then, covers the stream of material situations and happenings with all that they intimate in terms of emotions, values, thoughts.” Under this notion of the flow of life, filmic imagery can be more than evidence of something's existence. It can connote the subjects’ material reality, and occasionally even their inner world. The brave and confident expressions on the faces of Kim Chong-sŏk and his fellows tell us, at the very least, that even as political outlaws they did not want to appear meek before their executioner, who simply had a conflicting ideology. This potentiality of film to connote the material reality of open-ended life, with which Kracauer denied photography's affinitive ability to identify, can be related to one of my initial questions: “What can a piece of film footage bring to our time that a photograph cannot?” Film scholar Paula Amad (2010: 21) argues that film, with its counter-archival10 substances of “disorder, fragmentation, and contingency,” has “even more arbitrary and undisciplined style of evidence” than photography, so that it “undermines even further the archival support it inherited from photography,” such as the archive's physical data management systems like “the index card, filing cabinet, and the archive” itself. As in Giorgio Agamben's (2000: 55) famous statement, “The element of cinema is gesture and not image,” what makes film more arbitrary and undisciplined than photography is its gestural and movable feature. The moving images of bodies, although they are only fragments of a moment of time, present themselves as “gestures that have diverse latent potentialities of the bodies that would have had an infinite number of directions after the moment of time” (H. Kim 2020: 698).

Additionally, I want to draw attention to the elastic position in which Kracauer situates the viewer-interpreter of a camera image.11 The viewer-interpreter should presuppose a wide range of possibilities where the image depicts life, because the affinitive two—film and life—are both open-ended. I argue that this approach to camera images should be seen as a qualitative method. That is, viewer-interpreters are not decoding quantifiable or objectifiable data when they interpret the camera image but rather are facing a situation comparable to that of an ethnographer conducting fieldwork. They are deconstructing and recombining the “fabrics” of image and life, making “quilts,” in ways that simultaneously recall photography theorist John Tagg's (1988: 100) and qualitative methodologists Norman K. Denzin and Yvonna S. Lincoln's (2005: 4) accounts. The open-endedness of both image and life implicates “reference” that lies “behind” them—“a subtle web of discourse through which realism is enmeshed in a complex fabric of notions, representations, images, attitudes, gestures and modes of action which function as everyday know-how, ‘practical ideology,’ norms within and through which people live their relation to the world” (Tagg 1988: 100). The viewer-interpreter of the camera image, as a qualitative researcher, is “a maker of quilts” and “uses the aesthetic and material tools of his or her craft, deploying whatever strategies, methods, and empirical materials are at hand” (Denzin and Lincoln 2005: 4). The relative arbitrariness in interpreting camera images does not mean that the subsequently generated data is of lesser value but rather suggests that more “observations” and more “inscriptions” are necessary. In providing these, viewer-interpreters can turn what they see in camera images from “passing” imagery, which “exists only in its own moment of” display, into an “account, which exists in its inscriptions and can be reconsulted” (Geertz 1973: 9, 19).

Observing and inscribing the film footage of former “comfort women” also allow us to give an account of something that has thus far been left out of the discussions around the footage's evidentiality. The women's facial expressions in the footage released in 2017 can be interpreted as reflecting their nervousness and anxiety over being in front of the camera (see fig. 5). While the footage does not show all the relevant parties, it is obvious that there were American soldiers behind the camera, possibly one cause of the women's anxiety. It is also possible that these women, while under sexual slavery by Japanese soldiers, were taught that the US Army, the enemy of the Japanese Army, was evil and would not release them upon capture. Even though their oppressors had run away, these women's situation was fundamentally as prisoners of war, captured by their former enemies. Moreover, in this scene, another filmed party, the Chinese officers and soldiers—who went to the village to liberate it from Japanese military occupation and who, after this point in the footage, were going to release the women—wear expressions that seem to point to the actual situations the women encountered and would also encounter in their home village. Aware of what the women did for the Japanese Army, the soldiers’ smiles can be interpreted as somehow ill-natured. These images entail a range of different contextual fabrics, among them that the women were captured as POWs and were inspected by formerly demonized enemy soldiers; that the camera was set before them as a visualizing power, with many other eyes watching them being filmed; that even for the Chinese liberators, the women could still be sexualized; and that many former “comfort women” would return to their families and speak out about their experiences only to be labeled the shame of the family or even of the nation. Considering these fabrics, the Chinese male soldiers’ smiles disclose an affinity with the material realities of patriarchal, postcolonial South Korea, which the women would have had to confront first individually and then virtually in the present, as subjects of the camera image. When we think of the aforementioned, binary representation of “comfort women” in South Korean media—the “familialized” grandmother-activist and the idealized little girl—the reemergence of these young women, moving and alive in the film footage, carries a lot of weight for the South Korean mediascape. Their apparent fear and anxiety echo those of present-day women confronted by violence and cruelty as well as by the gaze of the media. Moreover, as the subject of camera images, they speak to present-day audiences through their recurring presence in exhibitions, wrenching open the “silencing structure” (Park 2005: 181), the structures that have made these women—gendered victims of violence—invisible to the postcolonial nation-state, its media, and its academia.

Returning to our initial discussion about NARA's archival principle and positionality, it is important to note that the researcher wanting to collect camera images from its archives should be aware of those images’ characteristics as qualitative data. Despite the long-standing myth of camera images’ indexical nature, there is plenty of room for different interpretations rooted in knowledge of the images’ creators, collectors, appraisers, and archivists, as well as in the viewer-researcher's own subjective positions. The researcher wanting to discover documents and reconstruct a narrative of the past, therefore, should always question the archive's logic and structure and allow for the possibility of disproving the archive's metadata. This methodology better positions us for postcolonial approaches to Western or former imperial archives.

This work was supported by the Ministry of Education of the Republic of Korea and the National Research Foundation of Korea (NRF-2019S1A5A8040231).

Notes

1

The research project began at the Human Rights Center at Seoul National University in 2014 and has since distinguished itself from the center, following the retirement of its chief researcher as the center's director. I temporarily joined the project's archival research and collection activities from May to June 2017 as a person in charge of conducting research on film materials at the US National Archives II in College Park, Maryland, and did not get involved in the two rounds of public releases of the collected footage in July 2017 and February 2018. Hereafter, I will refer to the team in charge of the project as “the research team.”

2

The term used in this article for victims of wartime sex slavery under the Imperial Japanese Army, “comfort women,” is in quotation marks to indicate that it was used by the Japanese military to designate such victims, underscoring the fact that the victims’ sexual service was positioned as a means of comforting soldiers and the fact that it was not a name chosen by the victims themselves but a historical title.

3

NARA (National Archives and Records Administration), “Research Room Rules,” February 6, 2019, https://www.archives.gov/research/research-room-rules.

4

In 2021, the person in question publicly revealed his identity. According to him, whose real name was Ch'a Pok-hwan, he had not even known about Chi Man-wŏn's claim that the person in the image was a North Korean spy until his wife by chance watched the documentary Kim-Gun on television and recognized her husband's sharp eyes from the photograph. In May of the following year, the May 18 Democratization Movement Truth Commission officially acknowledged that Ch'a was the real subject of the photographed citizen soldier who had been called “Kim(-gun),” overriding Chi's allegation (Hong 2022). While this conclusion appears to refute the documentary's final hypothesis that Kim might have been killed by the martial law troops, Kim-Gun's approach to answering Chi's accusation through the belief in optical evidentiality is worthy of notice, as discussed in the following part of this article. The director of this documentary commented that what his work had found as truth was “a vacuum,” not any facts about who might be the real Kim, even before Ch'a Pok-hwan turned up (C. Pae 2020: 22). Juyeon Bae (Pae Chu-yŏn) relates the self-consciousness of younger documentary makers who did not experience the violent era of the May 18 Democratization Movement to the issue of postmemory (22–23).

5

By the same token, this article does not include images from the footage The Battle of Tengchung, China (RG111-ADC-2417). It was one of the pieces of footage collected by the author in May and June 2017, and later that year, the research team confirmed that the footage contained images of the dead bodies of Korean “comfort women.” The film footage is accessible onsite at the US National Archives II in College Park, Maryland.

6

“G-3 Daily Diary,” 1944, General Correspondence, 1943–1945, RG493 Records of US Army Forces in the China-Burma-India Theaters of Operations, 1942–1947, Container 7, US National Archives and Records Administration.

7

“G-3 Daily Diary.”

8

This, to a lesser extent, is also relevant to the previously released images of moving “comfort women.” Nevertheless, this section focuses more on the direct representation of violence and violated bodies of victims.

9

Hŏ Yun (2018) historicizes the representation of “comfort women” in South Korean popular narratives. Until the 1990s, representation of the victims was rather frequently sexualized (136–39). However, since a 2004 controversy over a celebrity actress's nude photo collection that specifically used the theme of “comfort women,” the media have tended to adopt desexualized representations such as that of a little girl (139–42).

10

By “counter-archive,” Amad (2010: 21) means “a supplementary realm where the modern conditions of disorder, fragmentation, and contingency came to haunt the already unstable positivist utopia of order, synthesis, and totality.” According to her, film has a counter-archival value and presents “multifaceted challenge to historicism” (21).

11

I use the term “camera image” here to designate both film and photography. While it is important to note the potentialities of film as a medium of the flow of life, when we think of the means by which both archival photography and film footage are accessed as historical documents—the excavation of these materials from audiovisual archives under specific national, ethnic, racial, and/or geopolitical categories; the identification and verification of information believed to be conveyed by the materials; the release and exhibition of the materials to specific or general audiences to support historical claims—the search for the defining difference between the two media comes to carry less relevance to another question whether the subject can speak, especially in the current era of digital media.

References

Agamben, Giorgio.
2000
.
Means without End: Notes on Politics
. Translated by Binetti, Vincenzo and Casarino, Cesare.
Minneapolis
:
University of Minnesota Press
.
Amad, Paula.
2010
.
Counter-archive: Film, the Everyday, and Alber Kahn's Archives de la Planète
.
New York
:
Columbia University Press
.
Chŏng, Chin-sŏng, ed.
2018
.
Ilbon'gun ‘wianbu’ kwan'gye miguk charyo III
일본군 ‘위안부’ 관계 미국 자료 III (US Documents Relating to “Comfort Women” in the Japanese Army, vol.
3
).
Seoul
:
Sŏnin
.
Chŏng, Hŭi-jin.
2019
. Designated discussion at the conference “
War, Women, and Violence: Remembering the ‘Comfort Women’ Transnationally
,”
Sogang University, Seoul
,
South Korea
, March 8.
Chŏng, In-yŏng.
2003
. “
Yi kyŏng-mo ŭi tak'yument’ŏri
” 이경모의 다큐멘터리 (“Yi Kyŏng-mo's Documentary”).
Ŭmsŏng sinmun
음성신문, October 1.
Cumings, Bruce.
2002
. “
War Crimes and Historical Memory: The United Nations Occupation of North Korea in 1950
.” In
War and Cold War in American Foreign Policy, 1942–62
, edited by Carter, Dale and Clifton, Robin,
156
78
.
New York
:
Palgrave Macmillan
.
Denzin, Norman K., and Lincoln, Yvonna S., eds.
2005
.
The Sage Handbook of Qualitative Research
. 3rd ed.
Thousand Oaks, CA
:
SAGE
.
Geertz, Clifford.
1973
.
The Interpretation of Cultures
.
New York
:
Basic Books
.
Hausman, Jim, and Il-hwa, Chŏng.
1995
.
Hausŭman chŭngŏn: Han'guk taet'ongnyŏng ŭl umjigin migun taewi
하우스만 증언: 한국 대통령을 움직인 미군 대위 (Hausman's Testimony: An American Captain Who Moved South Korean Presidents).
Seoul
:
Han'guk munwŏn
.
Hŏ, Yun [Heo, Yoon].
2018
. “
Ilbon'gun ‘wianbu’ chaehyŏn kwa chinjŏngsŏng ŭi kon'gyŏng: Sonyŏ wa halmŏni p'yosang ŭl chungsim ŭro
” 일본군 ‘위안부’ 재현과 진정성의 곤경: 소녀와 할머니 표상을 중심으로 (“The Representation of the ‘Comfort Women’ of Japanese Military Sex Slavery and the Problem of Authenticity: Focusing on the Girl and Grandmother Images”).
Yŏsŏng kwa yŏksa
여성과 역사 (Women and History)
29
:
131
62
.
Hong, Haeng-gi.
2022
. “
5 · 18 ‘kimgun’ ŭn pukhan'gun anin ch'a pok-hwan . . . ‘chi man-wŏn, myŏngye hweson sagwa hara’
” 5 · 18 “김군”은 북한군 아닌 차복환 . . . “지만원, 명예훼손 사과하라” (“‘Kim’ from the May Eighteenth Movement Turns Out to Be Ch'a Pok-hwan, Not a North Korean Soldier . . . ‘Chi Man-wŏn Must Make an Apology’”).
Sŏul sinmun
서울신문, May 13, 8.
IEAS (Institute for East Asian Studies, Sungkonghoe University)
, ed.
2009
.
Naengjŏn Asia ŭi munhwa p'unggyŏng 2
냉전 아시아의 문화풍경 2 (The Cultural Landscape of Cold War Asia, vol.
2
).
Seoul
:
Hyŏnsil Munhwa
.
Im, Chong-myŏng [Im, Chong Myong].
2005
. “
Yŏsun sagŏn ŭi chaehyŏn kwa p'ongnyŏk
” 여순사건의 재현과 폭력 (“Violence in the Representation of the Yŏsun Incident”).
Han'guk kŭnhyŏndaesa yŏn'gu
한국 근현대사 연구 (Journal of Korean Modern and Contemporary History)
32
:
103
32
.
Im, Kyŏng-sŏk.
2002
. “
Ijŭl su ŏmnŭn saramdŭl—kang ta-ryŏng, chosŏn kongsandang ch'aegim pisŏ
” 잊을 수 없는 사람들—강달영, 조선공산당 책임비서 (“The Unforgettable—Kang Ta-ryŏng, the Chief Secretary of the Chosŏn Communist Party”).
Yŏksa pip'yŏng
역사비평 (Critical Review of History), February,
247
72
.
Kang, Sung Hyun.
2019
. “
The U.S. Army Photography and the ‘Seen Side’ and ‘Blind Side’ of the Japanese Military Comfort Women: The Still Pictures and Motion Pictures of the Korean Comfort Girls in Myitkyina, Sungshan, and Tengchung
.”
Korea Journal
59
, no.
2
:
144
76
.
Kim, Han Sang.
2018
. “
Piŏnŏjŏk chiljŏk charyo: Sich’ŏnggak charyo rŭl chungsim ŭro
” 비언어적 질적 자료: 시청각 자료를 중심으로 (“Nonverbal Qualitative Data: Focusing on Audio-Visual Documents”). In
Munhwa sahoehak ŭi kwanjŏm ŭro pon chiljŏk yŏn'gu pangbŏmnon
문화사회학의 관점으로 본 질적연구방법론 (Qualitative Research Methods from the Sociology of Culture), edited by Chong-nyŏl, Ch'oe, Sŏng-gyŏng, Kim, Kwi-ok, Kim, and Ŭn-jŏng, Kim,
286
316
.
Seoul
:
Humanist
.
Kim, Han Sang.
2020
. “
Palgyŏndoen p'ut'iji sok ŭi pagyŏngsim ŭn muŏt ŭl marhanŭn'ga (hogŭn marhaji mot'anŭn'ga)? Sajinjŏk saengjonja ŭi yŏnghwajŏk hyŏnjŏn kwa p'osŭt’ŭ/singmin ak'aibŭ ŭi naengjŏn chisikch'eje
” 발견된 푸티지 속의 박영심은 무엇을 말하는가 (혹은 말하지 못하는가)? 사진적 생존자의 영화적 현전과 포스트/식민 아카이브의 냉전 지식체제 (“What Does Pak Yong-sim in the Excavated Footage Tell You [Or Not]? The Cinematic Presence of the Photographic Survivor and the Cold War Knowledge Regime of Post/Colonial Archives”).
Munhak kwa yŏngsang
문학과 영상 (Journal of Literature and Film)
21
, no.
3
:
679
709
.
Kim, Jihoon.
2020
. “
The Uses of Found Footage and the ‘Archival Turn’ of Recent Korean Documentary
.”
Third Text
34
, no.
2
:
231
54
.
Kim, Pong-su.
2018
. “
Il kŭgu seryŏk e hyŏppak tanghanŭn wianbu yŏn'gujadŭl
” 일 극우 세력에 협박 당하는 위안부 연구자들 (“The ‘Comfort Women’ Researchers under Threat from the Japanese Far Right”).
Asia kyŏngje
아시아경제, March 13, 10.
Kim, Sang-sin [Kim, Sang-Shin].
2016
. “
Chenk'insŭn kwa swellenbŏgŭ ŭi kirokhak iron pigyoyŏn'gu
” 젠킨슨과 쉘렌버그의 기록학 이론 비교연구 (“A Comparative Study of the Archival Theories of Hilary Jenkinson and Theodore R. Schellenberg”).
Kirokhak yŏn'gu
기록학연구 (Korean Journal of Archival Studies)
50
:
61
95
.
Kim, Tae-sik.
1990
. “
Sŏp'yŏng: Yŏsun sagŏn kwa sajin ŭi yŏksasŏng
” 서평: 여순사건과 사진의 역사성 (“Review: The Yŏsu-Sunch’ŏn Incident and the Historicity of the Photograph”).
Yŏksa pip'yŏng
역사비평 (Critical Review of History), May, 350–58.
Kim, Tŭk-chung.
2001
. “
Yŏsun sagŏn ŭi chinsang kwa kukka t'erŏrijŭm
” 여순사건의 진상과 국가테러리즘 (“The Yŏsu-Sunch’ŏn Incident and James Hausman”). Proceedings of the Conference for the Fifty-Third Anniversary of the Yŏsu-Sunch’ŏn Incident. Edited by the Chŏlla Chwasuyŏng Culture Institute,
Yeosu National University
, October 19,
3
27
.
Kim, Yŏng-hŭi [Kim, Young-hee].
2018
. “
5.18 ŭi kiŏk sŏsa wa yŏsŏng ŭi moksori 5.18
” 의 기억 서사와 여성의 목소리 (“A Study of Women's Voice in Oral History Narratives on the Social Memory of State Violence: The May 18th Kwangju Democratic Movement”).
P'eminijŭm yŏn'gu
페미니즘 연구 (Issues in Feminism)
18
, no.
2
:
149
206
.
Kim, Yong-su.
2000
. “
Ilbon wianbu p'oro sajin sok yŏsŏng 1-myŏng pukhan saengjon
” 일본군 위안부 포로 사진 속 여성 1명 북한 생존 (“One of the Comfort Women POWs in the Photograph Is Alive in North Korea”).
Yŏnhap nyusŭ
연합뉴스, August 28.
Kracauer, Siegfried.
1960
.
Theory of Film
.
New York
:
Oxford University Press
.
Kwŏn-kim, Hyŏn-yŏng.
2019
. “
Ch'immuk ŭn mari toeŏtchiman mal ŭn ŭimi ka toeŏssŭlkka? Tŭtki ŭi kongdongch'e rŭl yebihamyŏ
” 침묵은 말이 되었지만 말은 의미가 되었을까? 듣기의 공동체를 예비하며 (“Silence Has Become Words, but Have Words Become a Meaning? In Preparation for a Community of Listening”). In
Chŏnjaeng, yŏsŏng, p'ongnyŏk: Ilbon'gun ‘wianbu’ rŭl t’ŭraensŭnaesyŏnŏl hage kiŏk hagi
전쟁, 여성, 폭력: 일본군 ‘위안부’를 트랜스내셔널하게 기억하기 (War, Women, and Violence: Remembering the “Comfort Women” Transnationally), by Yun, Hŏ, Kazue, Muta, Ichiro, Tomiyama, and Hyŏn-yŏng, Kwŏn-kim,
64
73
.
Seoul
:
CGSI EPUB
.
Morris-Suzuki, Tessa.
2005
.
The Past within Us: Media, Memory, History
.
New York
:
Verso Books
.
Nichols, Bill.
2016
.
Speaking Truths with Film: Evidence, Ethics, Politics in Documentary
.
Berkeley
:
University of California Press
.
Pae, Chu-yŏn [Bae, Juyeon].
2020
. “
P'osŭt’ŭmemori wa 5.18: Tak'yument’ŏri yŏnghwa ‘kimgun’ ŭl chungsimŭro
” 포스트메모리와 5.18: 다큐멘터리 영화 ‘김군’을 중심으로 (“Postmemory and the Gwangju Democratization Movement: Focused on the Documentary Kim-Gun”).
Sŏganginmunnonch'ong
서강인문논총 (Humanities Journal)
57
:
5
35
.
Pae, Myŏng-jae.
2019
. “
Yŏsun sagŏn chaesim chaep'an kongso kigak andwae
” 여순사건 재심 재판 공소 기각 안돼 (“The Application for a Retrial of a Case during the Yŏsu-Sunch’ŏn Incident Avoids Dismissal”).
Kyŏnghyang sinmun
경향신문, June 12.
Pak, To.
2017
. “
‘1800-myŏng chuginŭn te 3-il..’. han'guk iyagiimnida
” “1800명 죽이는데 3일..” 한국 이야기입니다 (“‘Only Three Days to Kill 1,800 People’ . . . It's a Story from South Korea”).
OhMyNews
오마이뉴스, November 7.
Pak, Ŭn-ha.
2017
. “
‘Konggam irwŏnaenŭn chinsil ŭi him’ . . . wianbu yŏn'gut'im 3in i chŏnhanŭn yŏngsang palgul p'ulssŭt'ori
” “공감 이뤄내는 진실의 힘” . . . 위안부 연구팀 3인이 전하는 영상 발굴 풀스토리 (“‘The Power of Truth That Gains Sympathy’ . . . a Full Story of the Comfort Women Footage Excavation Revealed by Three from the Research Team”).
Chugan kyŏnghyang
주간경향 (Weekly Kyŏnghyang), July
25
:
18
22
.
Park, Soyang.
2005
. “
Silence, Subaltern Speech, and the Intellectual in South Korea: The Politics of Emergent Speech in the Case of Former Sexual Slaves
.”
Journal for Cultural Research
9
, no.
2
:
169
206
.
Raines, Rebecca Robbins.
1999
.
Getting the Message Through: A Branch History of the U.S. Army Signal Corps
.
Washington, DC
:
Center of Military History, United States Army
.
Ranke, Leopold von.
1874
.
Geschichten der romanischen und germanischen Völker von 1494 bis 1514.
Leipzig
:
Duncker und Humblot
.
Sekula, Allan.
1986
. “
The Body and the Archive
.”
October
, no.
39
:
3
64
.
Signal: Journal of the Armed Forces Communications Association
.
1953
. “
Please Credit U.S. Army Photograph: An Album of Signal Corps Photographs That Record History from the Civil War through Korea
.” May–June, 25–29.
Stapleton, Richard.
1983–84
. “
Jenkinson and Schellenberg: A Comparison
.”
Archivaria
17
:
75
84
.
Struk, Janina.
2004
.
Photographing the Holocaust: Interpretations of the Evidence
.
London
:
Routledge
.
Sturken, Marita, and Cartwright, Lisa.
2001
.
Practices of Looking: An Introduction to Visual Culture
.
Oxford
:
Oxford University Press
.
Tagg, John.
1988
.
The Burden of Representation: Essays on Photographies and Histories
.
New York
:
Palgrave Macmillan
.
Yang, Ch'ang-hŭi.
2018
. “
Yŏsun sagŏn min'ganin haksal pujŏng kukpangbu ipchang, sajin·saryo ro panbak
” 여순사건 민간인 학살 부정 국방부 입장, 사진·사료로 반박 (“The Department of Defense's Denial of the Slaughter of Civilians during the Yŏsu-Sunch’ŏn Incident Refuted by Photographs and Historical Records”).
KBS News
,
May
16
.
Yi, Hŏn-mi [Lee, Hunmi].
2017
. “
Cheguk ŭi wianbu wa kiŏk ŭi chŏngch'ihak
” “제국의 위안부”와 기억의 정치학 (“Comfort Women of the Empire and the Politics of Memory”).
Kukche chŏngch'i nonch'ong
국제정치논총 (Korean Journal of International Relations)
57
, no.
2
:
329
66
.
Yi, Kyŏng-mo. (1989)
2010
.
Kyŏktonggi ŭi hyŏnjang: Yi kyŏng-mo sajinjip
격동의 현장: 이경모 사진집 (The Scenes of the Turbulent Era: Yi Kyŏng-mo Photo Collection).
Seoul
:
Nunbit
.
Yi, Myŏng-hŭi.
2018
. “
‘Ilbon'gun wianbu haksal’ kirok yŏngsang ch’ŏt konggae
” “일본군 위안부 학살” 기록 영상 첫 공개 (“The First Release of Moving Images Recording the ‘Mass Killing of Comfort Women by the Japanese Army.’”)
Kyŏnghyang sinmun
경향신문, February 28.
Yu, Ch’ŏng-nyŏl.
1947
. “
Okchung ŭi kyemyŏngsŏng
” 옥중의 계명성 (“The Crowing of the Cock from the Prison”).
Sinch’ŏnji
신천지, September 1.
Zelizer, Barbie.
1998
.
Remembering to Forget: Holocaust Memory through the Camera's Eye
.
Chicago
:
University of Chicago Press
.