A fortuitous encounter with testimonies

Hosted by the Women’s Human Rights Institute of Korea and the Research Institute on Japanese Military Sexual Slavery, and organized by the Humanities Korea (HK⁺) Research Center for Reconciliation and Coexistence in the Contact Zones of Chung-Ang University, Encountering Testimonies, an experimental exhibition of the AI-based interactive testimony content, was held from October 27 to November 7 at the Gallery MEME in Insa-dong. Although centered on interactive testimony content, the exhibition allowed us to get a glimpse of various concerns about testifying beyond technical performance and the place of encountering testimony after testimony. The English title of “Encountering Testimonies” demonstrates the character of the exhibition even better than the Korean counterpart, explaining the unexpected encounter with “testimonies” and the opportunity to meet multiple testimonies, not just one. Testimony here is neither giving nor listening, but encountering. The reason this encounter is possible is attributable to the fact that the testimony has already been given and thrown into the spot.

The introduction to the exhibition explains the meaning of testimony as follows: “Testimony refers to not only their memories, but also all of the “words” that reveal the shape, direction, and will of their current lives. Scarcely had these words been uttered when they left the witness and took on a life of their own and began to appear everywhere.” The testimony given is words that left the victim and no longer vest in the victim. As Ichiro Tomiyama^[1] pointed out, this is contradictory given that the realm of “testimony” is a place where someone refuses to let others speak for themselves. At this point, the predicament of the field of discourse surrounding testimony can be seen in the discrepancy between the possessive cases of the act of speaking and the words spoken. Nevertheless, in that the act of bearing witness inevitably summons the listener, the testimony becomes a collective speech act, not belonging to anyone. Primo Levi, who was a prominent Holocaust survivor, remarks that the warnings of Schutzstaffel (SS) soldiers of Nazi Germany about “no listeners” remain as persistent afterimages to survivors, connecting the issues of witness and listener. 

Simon Wiesenthal remembers that the SS soldiers showed pleasure, while cynically warning their prisoners: “(Omitted) Even if some evidence is found, and even if some of you survive, people will say what you are talking about is too terrible to believe. They’ll say it’s exaggerated propaganda of the Allies and believe us in denying everything. Rather than believing you. The history of Lager (concentration camps) will be written by us.” Oddly enough, these same thoughts (“Even if we talk, they won’t believe us.”) used to rise to the surface of the prisoners’ despair in the form of a nightmare.^[2]

The effectiveness of an utterance raises constant concerns about whether it will reach the listener. Testimonies always call for common utterances in that it is the speech acts with the listener in mind. In addition, as So Young-Hyun noted, considering the oral situation in which testimonies are given, it is clear that they are “collective voices created in the interaction of ‘speaking’ and ‘listening’.”^[3] That being so, how can these collective and common voices be reached by whom and to whom? 

Speaking in succession 

This exhibition is not an exhibition of imparting the details of the testimonies of the Japanese military “comfort women” survivors like a collection of testimonies but contemplating the context and acceptance of testimonies that have already been given when they are placed in different times and formats. It is mainly divided into four parts: the intro video; the painting and photo work by artist Jung Jungyeob; AI interactive content; and the media artwork by director Choi Kyung-jun. Spectators can change the order of viewing, but if there are not many visitors, moving along each space arranged in a straight line is the most ideal. 

Starting with the testimony of the late victim-survivor Kim Hak-sun, the intro video explains the process of testimony, the pain of testifying, and the reason why she had to bear witness. Despite the pain of giving testimony, victim-survivors continued to testify “to inform and remember,” and through this, they had experiences of “meeting supporters” and “healing little by little.” In this process, they emerged as “victims, activists, and witnesses” and “remained eternal witnesses.”^[4] At this point, testimony becomes an absolute being that defines victim-survivors. With regard to the absoluteness of these testimonies, Primo Levi goes so far as to say that the “lucky” survivors are those who survived to testify.^[5] And testimony includes not only words spoken but also non-verbal forms including the traces of violence engraved on the body (Jung Jungyeob, “Morning glories are prettier than cherry blossoms”) or crookedly written letters (Choi Kyung-jun’s media artwork). 

If so, can an unvoiced, unvisualized experience also become a testimony? Emily Jungmin Yoon’s collection of poems “A Cruelty Special to Our Species” and director Emmanuel Moonchil Park’s documentary “Comfort”, which were introduced as exhibition-related programs, take notice of these untold testimonies and ones that have been uttered but not visualized. Emily Jungmin Yoon’s work, named “found poetry,” reconstructs the testimonies of the “comfort women” survivors in the form of a poem, paying attention to the similarity of connotations, silences, and voids between the margins of the poems and testimonies. “Comfort” is also a kind of “found footage” based on the interview video of the late victim-survivor Kim Soon-ak, which was filmed during her living years. It sheds new light on life after Kim Soon-ak’s return, which was vocalized but deleted in this exhibition due to concerns about distortion. 

Unlike the “found” work in a general sense, however, the two works ask for a rereading of “testimony” as a kind of speaking and writing in succession by reintroducing the “original” into other contexts, rather than recombining the works they have discovered to derive new meanings. This is a characteristic found in other works introduced in this exhibition. The voiced testimony is brought back to life by repeating after the “testimony” (works of Choi Kyung-jun and Emmanuel Moonchil Park), by the viewer transcribing testimony (transcription table), and rearranging the uttered words through their own means of expression (works of Jung Jungyeob, Emily Jungmin Yoon). Aleida Assmann says to the listeners of the testimony, “the listener must be willing to share the testimony and become a co-witness or secondary witness of the memory that he or she helps to extend in space and time.”^[6] In the process of speaking, rewriting, and arranging the testimonies that have been given, the listener becomes a common witness, not a simple listener, through which the testimony evolves into a common voice. 

From listening to asking “again” 

Taking up the main space of the exhibition hall is the AI-based ​​interactive testimony content, a new type of exhibition project that uses AI technology to find answers to questions from the database of pre-recorded testimonies of two survivors, Lee Yong-soo and Lee Ok-sun. Initially, this project aimed at the continuation of testimony, preparing for an era in which the witnesses would be no longer alive. That’s why the first name of this project is “Eternal Testimony. However, as a wide variety of experiences of testimony tell us, the perpetuation of testimony verges on an impossible project in the sense that the testimony per se changes and its interpretation varies according to the political, social, and cultural context in which witnesses and testimonies are placed. In addition, there had existed concerns before the exhibition about whether the screen used as a means of communication may act as the screen that would obscure the vitality of testimony, and whether the limitations of technology and the “matching system” may make this project a kind of “enlightenment” project flattening testimonies and presenting only exemplary answers. These concerns turned into new anxiety and questions about the discursive field in which the testimonies were placed the moment I faced the exhibition project.  

The exhibition is divided into two screens, and the viewer faces Lee Yong-soo and Lee Ok-sun via each screen. The life-size ultra-high-resolution screen makes it clear that it is not an exaggeration to say “this exhibition seeks to explore the possibility of new communication through the experience of “directly” facing the testimonies of victims,” written in the exhibition introduction. In particular, when the character sitting with an expressionless face on the screen quickly answers “Hello” with a big smile in response to the visitor’s greeting, the borderline of the inhumanity of the new technology crumbles down, and even an unexpected intimacy is formed. 

The problem arises after that. After greeting, visitors come to be troubled by what to ask in what language. In order to alleviate this difficulty, example questions are prepared at the site of the exhibition, but for those who are familiar with their stories, these questions are rather standardized. My lack of imagination also kept me from creating new questions going beyond stereotypical questions. After thinking for a while, the question that came out was “Have you eaten?” I couldn’t choose but keep asking after her even after that question. At that moment, I reflected on whether I had forgotten the truth that listening includes questions in the process of thinking about the “community of the listener” and shifting the thinking of “from asking to listening,” which were raised after the fourth volume of the testimony collection and after Lee Yong-soo’s press conference. Perhaps this is where the validity of the questions posed by this tangible AI interactive testimony content to the viewer lies: bringing up issues again about what to ask, not what to listen to. That’s because asking well must be accompanied by the constant consideration of the questioner. This, of course, would be to restore asking within the process of listening, not a reconversion or return to asking. 

Additionally, the fact that the conversation with AI is made in an open place using a microphone poses another question. Whether the testimony of a witness is given in a private relationship or an open venue, the interlocutor’s question has been easily obscured as long as it is dealt with in the public sphere. While the questioner’s language is described neatly or the interlocutor steps back from the camera, his or her voice and face are erased. However, in this exhibition, questioners must reveal their faces in an open space and speak in their own languages. As many witnesses of traumatic events confess, testifying leaves pain in the witnesses. In addition to the pain of having to recall the incident by testifying, there is a problem of attention brought about by the action itself in the public sphere. Soojin Kim remarks the interviewer can also experience traumatic transference in a one-on-one interview.^[7] Of course, in this exhibition project, the screen “protects” the audience from this transfer experience. Nevertheless, viewers experience the weight and predicament when their language is uttered in a public site “with the face revealed.” 

Moreover, issues about the simplification of the aforementioned testimony content remain unresolved. Aleida Assmann says that what makes video testimony different from court testimony and autobiographical descriptions is that it takes on “a less elaborated form that leaves room for open-ended passages, such as pauses, periods of silence, uncompleted sentences, innuendo.”^[8] This does not take into account that video testimony can also be elaborated through editing, subtitles, sound correction, etc., but it is clear that video testimony excessively shows the gestures, facial expressions, and pauses of witnesses that language cannot reveal.^[9] However, the silence and gestures of witnesses are deleted or limited in AI content, and technical errors in place of silence, including not recognizing the interlocutor’s question or no answer matching the question, create silence for viewers, not witnesses. Although these technical errors can be improved to some extent, it is virtually impossible to prepare all possible question-and-answer videos. It is necessary to face the limitations of these AI testimony content, but these errors are not necessarily negative in that they reveal the place where the testimony content is produced. In the first place, the victim-survivors were not allowed to testify and see everything. The important thing is to consider the context and form in which the testimonies we face are placed and to think about how to create them into each montage. Only through this will we be able to imagine true listeners, common witnesses, and common voices.