Voices From S-21: Terror and History in Pol Pot's Secret Prison (18 page)

I was losing some of the people who were working with me. One day we were working together and then they were taken away. And they were killed. I felt anxious. I thought: “Today it’s their turn. I don’t know what will hap-pen tomorrow.”
³²

Everyone at S-21 was encouraged to be suspicious; as Him Huy told Peter Maguire, “We were all spying on each other.” Everyone was looking over everyone else’s shoulder and also looking in all directions. Fam-ily members were far away. Friendships provided little or no security, and patronage could be withdrawn at any moment. Every act could be construed as political. The nation was at war with its external enemies, society was at war with itself, the Party was at war with “hidden enemies,” and people were at war with their shortcomings. The Organization’s authority and legitimacy were unquestioned, but its principal message was that there were people concealed inside the Organization seeking to destroy it. By defi these “enemies” were simultaneously known and unknown, visible and unseen, outside and inside S-21. To cap things off, everyone might well be lying. A former DK cadre told Steve Heder in 1980, putting the situation rather mildly: “People were insecure psychologically [in the DK period]. People feared being wrong unconsciously or being fingered, [we] just kept smiling but [we] were tense inside.”
³³

The confessions that survive from S-21 vary in length, completeness, and interest. Roughly 4 percent of the microfi ones and an even smaller percentage of those in the DC–Cam archive are less than three pages long. Most of these were composed soon after the prisoners’ arrival, before any interrogation had taken place. They usually contain skeletal biographical data and mention no offenses. In several cases, the data sheets are annotated: “Of no interest. Discard”
(boh chaul).
Since the documents have survived, the “discard” orders may have referred

to the prisoners themselves; “discard” was one of the euphemisms used at S-21 for “kill.”
³⁴

While 2,013 of the microfilmed confessions, or roughly half the total of those in the microfilmed archive, were obtained after a single interrogation session, the remainder were composed in successive versions, those of cadres sometimes stretching over weeks or months. Most of the confessions in the microfilmed archive and at DC–Cam run between ten and forty pages. The prisoners I categorize as “cadres”— those who were over thirty years old and had revolutionary pseudonyms—often wrote confessions of several hundred pages.
³⁵

Over time, the format and style of the confession documents changed. Some of them were written by the prisoners themselves, and a few early ones were composed by prisoners and interrogators on alternating pages, but most of the handwritten confessions appear to have been transcribed by interrogators or document workers from tape recordings or notes. None of the tapes, which probably included questions as well as answers, has survived. Some confessions are obviously first drafts, with excisions and corrections by the prisoners, document workers, or senior cadres. In a few cases the drafts have survived alongside subsequent, typed versions, while other confessions, annotated “Don’t use” or “Don’t summarize,” exist in draft form and presumably never left the prison. Several confession texts from 1976 include questions written in an interrogator’s hand—even, on two occasions, Duch’s
³⁶
—followed by answers written by the prisoner. In nearly all cases, however, the questions and suggestions have disappeared.

By 1978 most confessions were typed. A number were prepared in multiple copies and stapled into booklets. We know from the Tuy-Pon notebook that by 1978 six copies of important confessions were normally prepared. Two of these were sent to “the Organization,” and the remaining copies were sent to the prisoner’s former work unit, the security office in the prisoner’s sector, and its counterpart in the relevant zone. The sixth copy was retained for the S-21 archive.
³⁷

Many of the early confessions at S-21 resemble prerevolutionary Cambodian police reports. Drawing on the French police tradition of the
procès verbale,
they recorded a prisoner’s initial, often self-incriminating declaration. Many of them include such colonial-era idio-syncrasies as spelling out dates, calling the prisoner “the named,” and so on. Most confessions at S-21 were authenticated by being signed or thumbprinted and dated by the prisoner on each page, another carryover from prerevolutionary police practice. When an interrogator wrote

out the confession, it was usually authenticated by the prisoner and countersigned by a document worker present at the interrogation. The elaborate format of the confessions and the files of which they formed a part suggest that Duch and his colleagues were proud of their thoroughness, modernity, and sophistication. They wanted S-21 to be considered a model interrogation center and saw themselves as professional security experts.
³⁸

We know very little about the way prisoners were processed for interrogation when they arrived or how interrogators were briefed from one day to the next, but forms relating to nineteen prisoners under interrogation in August, September, and December 1976, the only ones of their kind to come to light, provide a glimpse of these procedures.
³⁹
The headings on the forms asked for the prisoner’s name, pseudonym, work unit, and “strings”
(khsae),
the word used for patronage networks. Handwritten comments then set out the questions that had been asked on that day and suggested a line of questions for the next interrogation session. For several prisoners, forms survive from as many as six successive interrogation sessions. Interrogators recorded the prisoners’ health (fi were listed as “weak” and three as “normal”), indicated whether torture had been used (in these cases it had not), and noted the condition of each prisoner’s shackles and the key needed to open them. In one case, the interrogator observed that the prisoner had written ten pages of his confession the preceding day.
⁴⁰

By the end of 1976, most confession texts at S-21 were in a four-part format that endured with few alterations until the collapse of the regime. In the first part, prisoners provided their “life stories,” named their relatives and associates, and listed their work units. These curric-ula vitae were normally followed by a section titled “history of [my] treasonous activities” or “my political biography,” with data arranged in chronological order. A third section, called “plans,” described what the prisoners would have done had they not been arrested. Most confessions closed with lists of a prisoner’s associates, or “strings of traitors,” with indications of their whereabouts. In some cases, the “strings” included everyone, even dead people, who had been named in the confession.

The autobiographies were inspired by a peculiarly Communist genre of writing, the self-critical life story or
pravatt’rup.
⁴¹
Before 1975, autobiographical narratives had been rare in Cambodia, and the biographical genre itself enjoyed no particular status. In DK, on the other hand, as in other Communist countries, self-critical autobiographical narratives of Party members were repeatedly solicited, compared, and kept on file. The practice of writing autobiographies and their occasional use as heuristic texts seems to have been more important in the Chinese and Vietnamese Communist parties than elsewhere. The practice was widely used by the Chinese in the 1940s. David Apter and Tony Saich, writing about the phenomenon at that time, suggest that its goal was “exigeti-cal bonding.”
⁴²

In Cambodia, self-critical autobiographies were featured in the regular “livelihood” meetings. Many non-Communists encountered the genre for the first time when they were evacuated from the towns and cities in April 1975 and asked at roadblocks to compose them. When the procedure became a national routine in 1976, people were periodically asked to name their family members and associates, to describe their class origins, to list their political activities, and to set out their strengths and weaknesses. Finally, participants were made to suggest ways in which they might improve their own and others’ attitudes and behavior.
⁴³

The process of self-exposure, attested in many secret societies, was intended to purify the participants, reinforce the solidarity of the group, and display the Organization’s empathy and vigilance. For Party members, of course, the Party replaced the Organization. As Apter and Saich describe the experience of self-criticism in Maoist China, “One begins with sin and blemish, the purging removal of which is essential for enlightenment, not only for the self but for the collectivity, lest others become contaminated and polluted. Indeed, the notions of pollution and purification are endemic in rectification.”
⁴⁴
The biographies thus induced could also be used as incriminating evidence or to justify promotion. Everyone in DK was constantly on trial. A person’s trust-worthiness could evaporate if information in a
pravatt’rup
was found to be incomplete, misleading, “complicated”
(smok smanh),
or incorrect. Conversely, a “good” biography that included a “good” class background and praiseworthy activities could lead to Party membership, better work assignments, and enhanced personal security. Kok Sros has said that he owed his own rise within the Party (he became a “full-rights” member in 1978) to his hard work and his “good biography.” Nhem En, the S-21 photographer, has made a similar assertion.
⁴⁵
The autobiographies were measured against the Party’s requirements

of the moment, which is to say against the Party’s history up to the time when they were written and against the ever-altering tactical requirements of the revolution. Study sessions in DK repeatedly stressed the

importance of preparing “clean” biographical statements, and in early 1976 readers of
Tung Padevat
were warned against people with “systematically complicated biographies.” On the occasion later in the same year when he described the Party’s internal enemies as “germs,” Pol Pot told his subordinates that “life stories must be good and must conform to our requirements.”
⁴⁶

Senior cadres solicited biographies from their subordinates. In his confession, Non Suon dutifully regretted his incompetence in this respect:

My shortcomings included the fact that I did not follow up the biographical records of Party and core members in detail and then take measures to purge the Party. All I could see was the appearance of their actions. I was unable to grasp each of their essential origins, I failed to delve deeply into family roots. This provided the enemy with easy opportunities to penetrate and undermine the Party from within.
⁴⁷

By 1977, hundreds of thousands of ordinary Cambodians as well as Party members had written
pravatt’rup.
Thousands had written more than one. Over two hundred autobiographies prepared by S-21 personnel, for example, have survived in a booklet form that was apparently used throughout the country. These autobiographical booklets consist of questionnaires whose format so closely resembles that of the biographical sections of the confessions as to suggest that the questions in the booklets formed the basis of many interrogations.

The questionnaires open by asking about the subject’s name, revolutionary pseudonym if any, date and place of birth, sex, nationality or ethnic group, and marital status. The next item, “means of livelihood before entering the revolution,” was used to determine a person’s class status
(vannakpheap).
Most urban inhabitants of prerevolutionary Cambodia, known as “new people” after 1975, were lumped together as “royalists,” “capitalists,” or “petit bourgeois.” The class categories for rural inhabitants, on the other hand, took account of their material wealth, which was calculated on the basis of the dimensions of the land they owned, the materials of which the family house was constructed, the number of people in the family, and the quantity of livestock, ox carts, and farm equipment the family possessed. Suos Thi, the head of the documentation unit at S-21, for example, called himself a “middle-level middle peasant,” and Him Huy defined himself as a “lower-middle peasant.”
⁴⁸
In the confessions, references to a prisoner’s class origins are almost always missing, presumably because a “good” class background had been blacked out or a “bad” one confi by the prisoner’s treacherous actions.
⁴⁹

Before 1975, few if any Cambodians outside the clandestine Communist movement had engaged in class analysis. Instead, most of them distinguished broadly among the “haves”
(neak mean),
those with “enough”
(neak kuosom),
and the “poor”
(neak kroo),
groups roughly consonant with people who commanded
(neak prao),
“free people”
(neak chea),
and those who received commands
(neak bomrao).
⁵⁰