The China Initiative ended in 2022 as a result of accusations that the program prosecuted innocent people and used racial profiling against persons of Chinese ancestry. An often overlooked, but quite essential piece of the China Initiative was the FBI and DOJ’s attempts to frame innocent people. This occurred in some places by using manufactured cases, but also as a result of using a framing technique in linguistics called “semantic framing.” Federal prosecutors coming under immense pressure from Justice Main were given cases with very little solid evidence, but containing a lot of Chinglish, often of defendants’ own making. Many of these invented Chinglish phrases were nonsense or incoherent, but prosecutors nonetheless used the semantic framing technique to transform them into deadly evidence against even factually innocent defendants. When science-based linguistic validation techniques were used to debunk these false prosecutions, most China Initiative cases simply fell apart.
Semantics Framing
In the context of virtually any legal case, attorneys will use framing in order to interpret the words of a party in such a way as to support their client’s intention. Most people, even if very careful with their words, can have an entirely different and yet credible story told about them in the lawyer’s argument or presentation of evidence. The way it works is to put all of the context and circumstantial evidence related to the key phrases prior to and around the key phrase, such that it appears to be a natural extension of the corroborating evidence. Framing isn’t a huge problem in society because most people are in control of their words. But what if someone can make up things you said and attribute it to you—in legal evidence? That’s exactly what translators do, especially the translators working at the law enforcement agencies prosecuting the China Initiative cases.
Since the development of sociolinguistics research, linguistics expert has come to the conclusion that language itself has no meaning divorced from its social context. If a translator for the company issues a lot of nonsense phrases, it becomes doubly easy to frame the nonsense phrase in such a way as to put it into a context where it makes sense and supports a prosecutor’s position. Wrapping nonsense words in good social context rapidly leads to inferences that the word or phrase means something supported by the context, and can become an admission. There are traditional safeguards against this—law enforcement is required to disclose the “whole truth” in its cases. In order to go after hated enemies of the public, however, US law has developed several exceptions that allow expansive manipulation of the justice system.
Deleting the Whole Truth
Following 9/11, the United States developed an expansive law enforcement regime known as the Patriot Act. The Act greatly expanded law enforcement powers to respond to national security threats such as terrorism, and eroded a number of key civil rights such as Habeas Corpus. A good idea at the time was to require previously separated law enforcement and intelligence agencies to closely cooperate on cases. Clandestine officers would go abroad and collect information about terrorists in Afghanistan or Pakistan, then use the information to nab their conspirators in the United States. Lots of innocent people were kidnapped and incarcerated by the United States as part of the war on terror, and even US citizens were executed without a trial by using drone strikes. That kind of carelessness towards innocent people worked great when dealing with small, stateless terrorist organizations but rapidly got the United States into hot water when dealing with a sophisticated superpower. An intelligence-law enforcement cooperation defect exploited by the Patriot Act is not useful against a sophisticated adversary.
The defect in intelligence-law enforcement cooperation, has been made clear through controversy over the Kennedy assassination files. Historically, when President Kennedy was assassinated, the FBI was responsible for investigating the perpetrators. In practice, the CIA does a lot of overseas law enforcement investigations and is even the lead contact for many police agencies, as has been made popular in the UK TV show The Capture. The controversy with the Kennedy files is that the CIA deemed many of the relevant files classified and refused to share them with the FBI, therefore the FBI would never learn the whole truth of the case and it seems to have remained an unsolved mystery until today. Under law, the FBI is relatively transparent, but CIA operations enjoy absolute secrecy under its enabling statute, but the Church Hearings on the agency indicated that operations may be secret from anyone outside the agency, even the President.
We can only speculate what the spooks were doing because of the absolute secrecy, but based on the historical evidence I can surmise they were prosecuting innocent people who they felt threatened national security. As War on Terror civil rights erosion goes, this wasn’t a big deal at the time compared to things like extraordinary rendition and the plot of Zero Dark Thirty. The US already incarcerates vast numbers of blacks, Latinos, and now Muslims and Texas even has plantations full of black prisoners doing farm work, just like in the antebellum south. Being such a political non-problem, based on the China Initiative’s record, I would presume very early on that US intelligence services were providing incomplete truth to US law enforcement the whole time, withholding exonerating evidence and only showing incriminating documents. And when the FBI agents came to testify and swore to “tell the whole truth,” the federal government is lying to judges in order to crack down on terror. The individual agent isn’t lying, but he is an agent of the United States, which has classified exonerating information it won’t use.
In the context of semantic framing, providing an incomplete story does change the meaning of the words as a jury hears them. Thus, by telling an incomplete truth to a jury, prosecutors receiving incomplete information from intelligence services designed to maximize the chance of a guilty verdict, innocent people who seem dangerous and threatening can be incarcerated. Chief among these feared enemies threatening American national security is Huawei.
The Huawei Incident
When I researched the USA v. Meng Wanzhou case, the whole process of the case appeared to be one where a prosecutor was manipulated and pressured. Here is my speculation on what happened. US intelligence agencies were aware that through 2013, HSBC was deliberately laundering money and encouraging clients to use tricks to achieve just that. It knows because HSBC pled guilty to that. It looks like HSBC at the time solicited Meng to use a common money laundering technique—hire an unqualified translator to do KYC documents. Federal intelligence, knowing all about HSBC’s money laundering, cherry picked the Meng Wanzhou transaction, got the White House to authorize an operation, then the White House told the FBI and DOJ to prosecute Meng while pretending not to know about HSBC’s document laundering methods. However, we know the federal government was fully aware of HSBC’s document laundering, because they admitted to that knowledge in 2022 in a press release. Intelligence agencies gave the 2018 FBI/DOJ a tiny part of Meng’s story—conveniently omitting that HSBC was running organized crime at that time—and framed the narrative about “ordinary business cooperation” as one of defrauding a bank in order to violate sanctions. The way framing worked here was, the DOJ took a nonsense phrase and repeatedly stated in the case that it referred to an unaffiliated third party. The underlying Chinese means something totally different, but the fact of a Chinese source document existing and that the Hongkong Shanghai Banking Corporation knows a thing or two about the Chinese language anyway, were totally hidden and shielded behind the Patriot Act’s provisions for intelligence agencies.
China, for its part, was very displeased and saw the prosecution as an intelligence operation’s plan to take hostages. What the Chinese media analysis didn’t seem to recognize, is that there are quite a lot of us American legal professionals who object to black ops manipulation of the legal system and would have opposed such behavior if confronted on the points of the law. That includes the prosecutor whose name is on USA v. Meng Wanzhou, who later battled President Trump’s subversion of the legal election process and said “But I’m telling you what’s going to happen. You’re going to lose your entire department leadership. Every single AAG will walk out on you. Your entire department leadership will walk out within hours.” This is someone who sacrificed his career. After hearing about the linguistic analysis issue, the DOJ did indeed agree to release Meng Wanzhou. Overall, around 70% of the China initiative cases failed due to poor evidence, whereas DOJ prosecutions are supposed to fail 1% of the time or less.
Cases with much worse evidence in translation than Meng Wanzhou’s were introduced, for example, the Anming Hu case involved the FBI manufacturing false evidence. One of the translations in the case record featuring prominently in news reports is of “spiritual hurts”, which actually refers to psychological trauma and has nothing to do with religion or spiritual life. A case brought against a State Department employee alleges a handler called “the teacher”, a mistranslation so obvious even the lowliest of Netflix subtitle translators don’t make the mistake. The translations being used by the FBI to make criminal complaints during the China Initiative were absolutely astounding for their incompetence.
The reason the justice system overall failed in my opinion is that exposing the false evidence used in these prosecutions, once the falsification method from the Patriot Act was well understood, would be relatively easy. The defense needs merely involve a very highly qualified legal translator ideally with an ILR4 competency level and law background, to clear up the truth around the case. The less-qualified linguists will generally listen to and agree with a highly qualified linguist, in part because linguistics involves a lot of science and scientifically proving a correct answer is not difficult. It’s like asking a pharmacist to provide proof that drinking bleach won’t cause cancer, scientifically, this is very well established. That reality also creates a huge weakness for the FBI and DOJ when dealing with China: the country being targeted is scientifically and technologically sophisticated, enough so that sham translation evidence can be easily defeated.
Fixing the Issue
As indicated above, the root cause of the false prosecutions during the China Initiative is the FBI rule permitting substandard translations. The federal process causing it is not hard to identify; the FBI requirements state that merely having ILR 2+ competence in the second language is sufficient, but the ILR standards themselves state that work done with that proficiency level will lead to a “misunderstanding.” Moreover, law firms that turn to big translation companies for a second opinion, will also be given ILR 2+ or below language quality not just as to translation but also as to basic language skills. Very few translators have a language skill at the ILR3 or above level, in Chinese to English translation performance it’s several dozen out of hundreds of thousands that have a proficiency credential. For a typical learner, advancing from ILR2+ to ILR3 in Mandarin typically takes a master’s degree worth of additional education, according to writings of Professor Kubler’s assessment of the John Hopkins Nanjing center. Moreover, ILR3 itself is actually very minimal. The federal government actually has additional standards above 2+, for a major foreign military attack, ILR3 or higher is used, for existential threats to the United States, ILR4 or higher is used. To put this in context, the 9/11 attack only justified ILR3, a planned nuclear attack by Russia would be ILR4, because at ILR4 that’s where mistakes go away.
Federal policymakers pursuing the War on Terror seem to have permitted low competency standards in order to achieve lynchings on the cheap. If you look at the authorized pay scales for the ILR2+ level it runs around $45,000 per year whereas the very skilled ILR4 level translators can get starting salary compensation packages around $160,000 per year. Somewhere in the middle, the government would need to offer around $90,000 per year to get genuine ILR3 competency level. I recall a DOD linguist posting on internet forums (probably inappropriately!) I researched that apparently the DOD had ‘very few’ even ILR3 level people and that getting that skill is an achievement. The above indicates that boatloads of money can be saved by using inexpensive, cheap translators. Indeed, just about every country seems to have done the exact same thing with linguistics, treating it as unimportant primarily because foreigners are universally discriminated against. Indeed, during the 1970s, many Chinese-English linguists in both the USA and China were targets of harassment and violence for even learning the skill. A step down from terrorism is discrimination, which is the atmosphere in which language work is done, and by no means is Chinese-English translation treated equally vis a vis its importance.
Consider for example how Duke University professors discouraged people from speaking Chinese on campus, which is a violation of civil rights law. At the same time, Duke University has a Chinese language program and Shanghai exchange school which encourages future China experts to connect with those Chinese students on campus. How is that going to happen if your professors are bullying people into silence?
If Washington wants to have meaningful, useful China-related capabilities then ironically that means treating Chinese culture and people with full equality. In part, that means raising its language competency target to ILR4 and at least IL3, with financial incentives, because if the federal policymakers require ILR4 when their own lives are at stake from existential threats, why shouldn’t they do that when some foreigner’s life is at stake during one of their China Initiative prosecutions? As General McChrystal pointed out, recklessly locking up, torturing, and killing “foreigners” around the world only makes us less safe—that’s not his personal opinion, that’s literally the entire point of On War by von Clausewitz. Falsely prosecuting lots of suspected saboteurs creates an army of new saboteurs. Only by valuing Chinese language and culture psychologically and economically, can policymakers avoid arresting and prosecuting innocent people and narrow prosecutions on people who are actually guilty.
