Translators of Chinese have famously low motivation to their career, resulting in widespread incompetence and low pay. The lack of competence among such translators has also resulted in the lowest translation rates anywhere in the world, about six times lower than typical rates for European languages such as German or French. The fundamental reason is Chinese translators generally have very poor motivation to learn strong English or effective translation skills, and do not hold themselves to a high standard, unlike European translators. As a result, Chinese translation projects get very poor results, and generate very little revenue that could be paid to those translators. Thus, the best Chinese translators earn 50 times entry-level translators, whereas the same spread in France is about 6-to-1. Poor motivation, not market fundamentals, is the principal driver of the famous poverty of Chinese translators.
Employment Opportunity
A chief reason for the extremely low level of motivation among Chinese translators is the extremely poor employment opportunities in their field. When I talked to a number of translation students at several different translation programs around China, Europe, and the United States, virtually all of them indicated a desire to exit the translation profession before graduation. At the very best schools in the United States, I observed translation students as being more likely to take employment as translation project managers at large translation companies. This means that for law firms and corporate clients, the ultimate decision-maker on major business matters is often a 25-year-old fresh university graduate with no business experience and even, quite commonly, an intern. While not doing translation in itself, despite being trained primarily for translation, they are at least involved in a translation-related field.
The situation in China is a bit different because there is not just a handful of translation institutes as seen in the US/UK, such as MIIS and NYU in the United States, but rather the country has launched over one hundred graduate translation programs. Students complain of translation company rates having fallen to the equivalent of $1.5 USD/hour (30RMB/1000 characters), but if fraud is used, then the translation rate can increase to $4.5 USD/hour. The specific fraudulent technique is called quality fade, which is well-documented elsewhere and involves providing good quality initially, then an increased percentage of machine translations in subsequent work.
For translation clients, this creates a situation where a law firm for a corporation doing M&A due diligence orders 20 million words of Chinese to be translated at $4 million USD ($0.20/character) but gets back almost pure machine translation, also totally incorrect and perhaps even with some haphazard mass replacements. The project is outsourced to a company in Bulgaria ($0.1/character) for $2 million USD. The Bulgarian company outsources to a Beijing company for $1 million ($0.05/character), and that company outsources to a Wuhan translation company for $300,000 (less than $0.02/character). Almost all big translation companies talk about a “supply chain” despite, in theory, providing services: this is basically how the supply chain reaches through a network into China. Generally, no person in the supply chain is fully aware of who all of the stakeholders are.
The market equilibrium rate is generally set at the equilibrium of superficial accuracy — the very minimum the translator can do without being outright accused of faking the work. This generally involves massive copying and pasting into machine translation programs and, when doubts are raised about the quality of the work, providing explanations and justifications as to why the quality is good while hiding the fact that it is just a machine output. A number of international banks and law firms with China offices also, for the sake of cost cutting, have their own internal staff doing the exact same thing while deceiving senior management about the method in which the work was faked in order to reduce costs. As a result, Chinese business and legal translators actually experience worsening translation skills over time because the translation ability needed to gain certification is not actually being demanded by the market. Instead, they become experts in con artistry, forced to find ways to persuade clients that something copied out of Google Translate is a high-quality, artisanal product. Translation service quality is so bad that the professions of localization and interpreting have literally split off from translation.
As a result of the immense amount of fakery, Chinese translators are often presented with a choice: fake the work or go hungry. When a company threatens its employees with starvation if they do not cheat, most employees choose to cheat. These perverse incentives create a situation for Chinese translators where their skill and quality of work is totally irrelevant; the only skill that matters to them is how to use confidence tricks on clients in order to pass a machine’s work off as the real thing. This is by no means unique to China, as the central maxim of the book The General Theory of the Translation Company points out, “quality is not important.” However, all translation companies insist their quality is excellent, while most business models revolve around minimizing the cost of the translation itself and, therefore, the quality. The General Theory urges companies to emphasize “services” not related to translation, and, in this case, companies have responded by spending 97% of costs on supply chain administration.
Don’t get me wrong; having such an extensive supply chain that any corporate bank can fill any translation-related request of any size, any language, and any specialization within an hour – a nimbleness not matched in any other service industry except software apps – but this kind of positioning results in translation service revolving around fakery, notably in China but also reportedly in Eastern Europe. An industry revolving around reliance on client ignorance has extremely damaging results that I can see in China (and likely other countries – please let me know your experience!). In China specifically, before English cram schools were shut down by executive order, almost all translation students intended to be English teachers. This may have made some logical sense because English instruction in China still largely revolves around translation, a practice that has only been phased out at the university level. Students in legal translation programs say they are planning to become paralegals, not translators, due to inferior compensation paid to translators. Clients obviously value translation, or they would not pay things like the $4 million fee for a very large project, but what is the value of translation to clients?
Chinese Business Translation has Value
I have followed a number of companies and public policy initiatives over the years and have found that if a particular initiative succeeds without translation 99% of the time, then its success rate drops to 30% with translation. Of particular note is the China Initiative, where some cases were translated by government translators with spotty qualifications, and others involved privatized court translations. Among the government-translated cases, 70% ended in failure against the DOJ’s benchmark of failing less than 1% of the time when charges are brought. Another useful statistic comes from Netflix: 70% of subtitled television viewership ends because the subtitles are translated poorly. The historic record of the “go out” movement for Chinese corporations is marred by problems such as the Tencent failed North American launch, Trump’s attempt to ban ByteDance, and the actual shutdown of ZTE Corporation. During this time, China also literally banned blue cheese from import over translation issues. China bloggers urge woe to businesses entering China, for they will only experience doom and despair there.
The Huawei cases, specifically USA v. Meng Wanzhou, are a great case study of this particular issue. A business executive spent three years under house arrest because of a document that the plea bargain revealed was simply mistranslated, and government agencies typically assign liability to a client for a translator’s mistakes. Therefore, if Meng hired a translation company with underpaid translators, the fact that those translators felt a need to cheat to put food on the table is criminal conduct attributable to Meng Wanzhou. When I filed FOIA requests to the SEC about financial fraud, they said that they would indeed charge criminal fraud in this particular type of fact pattern. The general public does, as well. For example, Facebook’s and Apple’s robot machine translator algorithm has made numerous offensive mistakes over the years that have been directly attributed to these companies as intentional and offensive, even though it was well known this was just software.
Moreover, organizations tend to blindly rely on translations even if the translation is obviously incorrect. For example, Israeli police arrested and detained an Arab man over a social media post mistranslated by a machine translation program. This is unlikely to happen again for them, but when a translation labeled as a human translation is actually an incorrect machine translation in disguise —the necessary result of current Chinese translation service supply chain management —organizations are likely to blindly rely on it and, even if challenged, the translator is likely to fight back. The organization itself fights back because it feels threatened that its talent-sourcing capabilities are under attack. For example, FBI translator Sibel Edmonds pointed out that when she raised concerns about the translation unit’s reliability, the organization told her to drop the matter. A few years later, the FBI admitted the same translation unit bungled translation work that could have prevented 9/11.
Despite the huge value of Chinese translations to clients and society as a whole, young professionals are increasingly mentally checking out from the profession and looking for an exit. This is obviously very dangerous for relations between the US, UK, and China and bears at least part of the blame for the escalation in an era of increasing social and technical complexity. What can be done?
I might recommend, in general, making the translation process less dependent upon laissez-faire supply chains that lack transparency. Chinese translators are particularly demotivated because they are being treated like commodities and not like professionals, and this lack of motivation is leading to a widespread focus on con artistry, often described as “Chinglish.” Organizations, therefore, need to treat translators like professionals who are held to professional standards, or otherwise will simply be given machine translation dressed up to look like a human did it, which can be very dangerous. Industry regulation is also needed; unlike medicine, law, or even children’s toys, translation is also totally unregulated. In the US/UK, this would imply a mandate for the use of professional certification, and in China probably involve minimum pricing requirements to help drive quality-focused competition during the provision of translation services.
