Geoffrey Cain - The Perfect Police State

From August 2017 to September 2020, I interviewed 168 Uyghur refugees, technology workers, government officials, researchers, academics, activists, and a former Chinese spy who was preparing to defect. Most requested that I use pseudonyms if I published their interviews.

This was the only way I could write an accurate book on the Uyghurs, Kazakhs, and other ethnic groups that are oppressed in western China. Their families, who are still in China, can easily be harassed or tortured by the police or intelligence services. Some interviewees, public figures whose stories have already been widely published, requested that I use their real names.

The protagonists name in this book, Maysem, is a pseudonym for a young Uyghur woman I met in Ankara, the capital of Turkey, in October 2018. I interviewed her fourteen times from October 2018 to February 2021.

As I researched this book, I questioned myself about the wisdom of interviewing refugees who, under the cover of fake names, could take license to stretch their stories, or change their accounts to cope with unbearable trauma.

In fact, there is a long tradition among journalists and scholars of relying on refugee testimony when documenting political upheavals and human rights atrocities, sometimes withholding their names and identifying information. Researchers and writers have long relied on the testimony of White Russian, Jewish, Cuban, Chinese, and North Korean refugees.

Like some of these other writers, I spent my career as a foreign correspondent interviewing refugees and people in sensitive situations, whose careers or lives were at risk. I knew how to get around the downsides.

I rigorously cross-checked the statements of key characters with people who knew them and had heard their stories before. I also cross-checked my interviewees statements with other refugee accounts, media reports, the chronology of events on the public record, the Uyghur victim database at shahit.biz, the speeches in state media of Chinese leaders, human rights reports from Human Rights Watch, the Uyghur Human Rights Project, the US Congress and US State Department, and the work of a small group of researchers who analyze satellite imagery, computer data, and Chinese corporate reports.

For key interviews, I spoke with each person in anywhere from four to twenty-five sessions, spending between eight and sixty hours with each one, revisiting the same questions, and listening carefully for inconsistencies or omissions. In 2 out of 168 interviews, I found the stories changed in minor ways. In one other interview, I found that the interviewee withheld information that reflected poorly on him; he was attempting to convince me he was a victim and not a perpetrator.

I didnt use these three interviews.

In all other interviews, the stories remained consistent down to fine details, in dates, places, addresses, and names. Most of these details were verifiable on Google Maps satellite imagery, in leaked Chinese state documents and online tender documents the interviewees provided to me, in annual reports published online by Chinese corporations, and in my own observations and travels inside Xinjiang.

From December 2020 to February 2021, I worked with a rigorous fact-checker, Wen-yee Lee, a Taiwanese technology journalist and native Mandarin Chinese speaker, to double-check the material in this book. She reviewed interview recordings and transcripts, called the interviewees and double-checked their stories, listened to their stories an additional time for changes or omissions, and rigorously checked the original Mandarin Chinese sources cited and translated in English-language press reports and research. I further checked passages on the mechanics of artificial intelligence, facial recognition, semiconductors, DNA collection, and other technical fields with academic experts. These experts, some well known in their fields, vouched for the technological validity of what Uyghurs witnessed and told me about Chinas surveillance state. I take responsibility for any remaining errors.

Since Chinas surveillance state is growing in sophistication, I withheld or obscured some pieces of identifying information, such as the exact chronology, locations, ages, and the amount of time that transpired between events. When writing the story of the protagonist Maysemthe pseudonym is also the name of a historical Uyghur heroineI removed identifying dates and places, and obscured the amount of time that passed between events during her imprisonment. This was the only way to protect my sources while remaining true to their stories. If Id revealed too much, the intelligence services would have no problem tracking down whom I spoke with, along with their families and friends.

Uyghur names consist of the individuals name first and the family name second. Han Chinese names follow the reverse standard of family name first, individuals name second. Out of respect for the people who appear in this book, Ive written their names in English as they prefer to write it themselves. Some prefer to write their Han Chinese names in English with the given name first and family name second (Kai-fu Lee), as opposed to the usual convention (Xi Jinping). When writing Han Chinese words and names in our alphabet, the pinyin system, the international standard for the romanization of Standard Chinese, is used throughout.