Clients seeking legal services in China can generally expect to either receive counsel in broken English or advice copied directly from an AI assistant, a phenomenon particularly prominent in countries with languages very different from English. In either case, the results are low quality, unreliable legal counsel. While expert researchers recognize that any sort of language translation is highly specialized and technical, lawyers called on to serve clients typically rely on a single college course in legal English. A practical solution is to have legal linguists with the appropriate expertise bridge the gap. This article describes the problem and how that approach can work.
The Status Quo—ESL
Traditionally, professional translation has not been used at all for business operations, in part because translation quality has been so low that any use of translators in operations would cause the business to become dysfunctional. For example, the Harvard Business Review once opined correctly that all CEOs must learn good English because the lack of qualified business translators could make otherwise simple sales calls impossible.
This was also my experience working at a big four audit firm and later at a large law firm; lawyers and auditors simply did not trust the translators. Starting out as a lawyer, I was shocked at the attitude of many translators, who generally believed that their job was to simply make up nonsense words and put them into a sentence, regardless of whether the intended reader could understand them.
For a professional services firm of any type, this means that all bilingual work must be done by non-translators. A Big Four firm, for instance, would have accountants translate statements from the China tax office for M&A transactions. Without training, these statements were typically word-for-word, grossly distorted, or meaningless. New York office associates receiving these documents would simply exclaim, “I can’t work with this!” and file them away, never to be touched again.
For transactional documents, lawyers often “translate” contracts from English to Chinese and, in an effort to maintain fidelity under the outdated Xin Da Ya theory, grossly misconstrue the client’s intentions and create misunderstandings. A distinctive feature is how these contracts are translated word for word into Chinese, so the result looks nothing like a real Chinese contract, and the precedential framework needed to interpret it is destroyed.
With the large amount of bilingual work being done by lawyers, you might expect there to be an educational system in place to train professionals. Actually, according to local law professors, fewer than 1 in 100 students at international law-focused universities in China have learned English beyond the high school level. According to lawyers moving on to law firms and being asked to do English-language work, they are plunged into relying on a mixture of dictionaries and AI to handle all client work, and try hard to avoid clients noticing their lack of skill. Memorizing talking points before conference calls is a common tactic employed by these lawyers to save face. Nonetheless, the outcome is still superior to hiring translation companies that send totally unedited Google Translate back for everything.
The situation has not gone unnoticed by the Chinese government, and law enforcement and the exodus of foreign firms are highly correlated. In response, a number of conferences and seminars held across the country over the past few years have sought to solve the “legal English problem”. An emerging conclusion is that lawyers really don’t want to spend time learning English, but they really do want to collect legal fees. This is highly apparent in litigation when business deals collapse due to misunderstandings created by lawyers.
Incompetent Consultants
Into this void have stepped a number of business consulting companies capable of explaining the basics of China’s legal and business systems effectively. However, they are not actually experts in the field and depend heavily on AI chatbots to give advice, offering the same kind of insights you could get from a $30 tax filing software package or online business incorporation service. In one recent case, an online retailer relying on one of these services to export to China was fined and blocked over data privacy law violations.
While these consultants may appear to be very helpful initially, businesses are increasingly finding that they provide bad advice or make serious mistakes that prove to be damaging and costly.
Law firms have also copied this trend to some extent, setting up international law departments where English-speaking generalists serve clients akin to corporate counsel. Meanwhile, local Chinese clients are served by a different group of lawyers who charge far less but have stronger skills owing to specialization. These generalist lawyers typically do not consult with their own firm’s specialists and instead give clients relatively superficial advice from law practice databases. Even so, a casual perusal of even the most prestigious law firms’ websites reveals that they are not able to convey their legal advice very clearly, especially in comparison to the consultants. Often, it’s incoherent.
An identifiable root cause that can explain this situation, as related by the universities, is that nobody studying fields like law, tax audit, technology, or data privacy really wants to learn English. Even those who studied English as undergraduates often choose to go to law school specifically because they don’t like the language. As a result, businesses either get incompetent counsel or advice mangled by AI.
Lawyers are increasingly admitting to copying and pasting from ChatGPT to appear competent in front of their clients. Clients are therefore paying for the perception of reliable advice, unaware that AI-generated work is being passed off as legal advice just because it sounds smooth and coherent, and oblivious to the problems it could cause down the line.
Where Are the Translators?
The translation field exists largely to solve these problems. If we go back and look at history, treaties, the earliest legal agreements, were mediated by translators. Later on, however, the wheels seem to have somehow fallen off the cart during the rise of globalization, which resulted in a kind of split where lawyers began doing this work in-house. This is, in fact, how I got started with translation: outside translation companies were simply too unreliable to trust with important work. Indeed, law firms in China and elsewhere also have translation departments, yet even they are not trusted with important work.
In China’s case, translation associations have noted that translation companies have unrealistically low rates, so the industry has long been a front for AI output passed off as human work. Paradoxically, law firm translation departments are even worse, charging a fraction of the already low translation company rates for internal client matters. These rates are so low that department managers urge their subordinates to aggressively pass AI off as real work. As a result, the translation quality is not fit for any meaningful purpose, and the lawyers end up doing important translation work themselves, billing up to twenty times what their translator colleagues charge.
When looking at why the collaboration model used in these contexts is not producing anything of value, a good study on why this model fails is Legal Translation Outsourced by Juliette Scott. Practicing lawyers have zero interest in legal translation, and the information shared when work is delegated to translators is limited to a mere purchase order number. Since ISO defines quality as the ability to achieve a purpose, it is impossible for translators to provide actual quality when no purpose is stated. Furthermore, they are under a lot of pressure to cut costs, and these end up being cut directly from mission-critical content for lack of proper guidance.
As a result, professionals in business fields often find both outsourced and their own translators inadequate and resort to taking a DIY approach, mastering English (or pretending to) in order to provide services to international clients, with frustrating and disappointing results.
Translation Can Fix ESL Law Failures
The prior failure of lawyer-translator collaborations during the emergence of globalization can largely be traced to communication failures and can be fixed by targeting that weak link. Looking back at Scott’s research, we can see that international legal work is being organized in a zero-communication environment. That is, lawyers and translators don’t tell each other anything and simply pass files back and forth. Scott recommends using a questionnaire to solve the issue, but this has proven unrealistic and has never been adopted.
Instead, I believe the solution is to correct organizational defects through language brokering, which is characterized as “the informal act of translation by children and young people between a family member and a dominant language speaker.” In almost any law firm environment, a particular attorney will have to quarterback a deal, coordinating between a large number of professionals. In practice, this means that a bilingual attorney is the point of contact for an English-speaking client, and therefore, the attorney, through language brokering, is now responsible for doing translation. As a language broker, the attorney then delegates translation work to a translator, at which point the zero-communication environment sets in because the attorney is an incompetent translator, thus forcing their incompetence onto the translator team.
Occasionally, a business executive based in the US, Europe, or sometimes Singapore is put in the language broker role, typically someone very bilingual who has gotten bad results from attorneys. Yet these language brokers are doing the same exact thing as those attorneys. Their behavior doesn’t achieve any better results because they lack translation expertise and don’t know what information someone needs to translate a document, so they make the same mistakes.
A cure to this defect is to replace lawyers and executives serving as language brokers with qualified legal translators, who themselves usually hold legal qualifications. The result of qualified translators performing language brokering in this situation is that the context of the legal matter and the needs of the client are actually understood. For example, suppose a client has a question about structuring data processing agreements; a translator-broker will see the client’s question and proposed business model, and work with attorneys on providing advice that fulfills their needs.
On the other hand, if a company attorney or an executive sends only a file and a due date, the resulting memorandum will appear to have largely just fallen from the sky. Since semantic structures in Mandarin are very different from English, a translator given no context will probably translate the document word-for-word, producing something grossly misleading and not responsive to the client’s question. As a result, an informal translator (attorney) with context would do better and, in practice, is what tends to happen.
Can Language Brokering be Functional and Provide Value?
You might be wondering why translator-to-attorney delegation succeeds where the reverse fails. A key reason is that major professional specialties—medicine, legal, tax accounting, human resources, and data privacy—are highly regulated with tightly defined requirements. A standard tax questionnaire or business acquisition due diligence checklist can be easily found online and the need-to-know is well understood, as are the evidence-based procedures nurses in hospitals are required to follow. Without access to this information, providing even minimal service is completely impossible. With a translation, on the other hand, it’s possible to do at least a superficially passable job by receiving documents alone, without any other information.
There are also questions about the benefit that can be provided by translators this way, especially since clients have been happily purchasing these services for several decades now. An obvious benefit for clients I can see is cost. In China’s markets, fees charged by bilingual professionals tended to track what would be charged in the United States. For example, bilingual corporate law attorneys in Shanghai often charge $1,500 per hour, but having spent many years in the United States, they have much less China-law expertise than local peers. Lawyers in China, however, earn (and charge) on average seven to ten times less than New York lawyers, meaning clients pay ten times the cost for half the lawyer.
These professionals usually spend a lot of time overseas to learn good enough English, something beyond the reach of most people. In many cases, these tend to be people with wealthy families paying upwards of $200,000 for their foreign education and some experience; thus, the main limiter is not talent, it’s inheritance.
Translators serving as language brokers, however, can help businesses overcome this scarcity and get superior talent at a fraction of the cost. In fact, this kind of service has been around for at least ten years, though a major barrier, solved in more recent times by AI advances, has been the enormous cost of translation relative to legal services.
Challenges in Getting Language Brokering to Function
Some major challenges can arise for organizations that attempt to use language brokering. As has been pointed out by translation commentators, a typical translator is someone who has a lot of interest in learning languages but no interest in anything else, especially technology or business. On internet forums like Quora and Reddit, we can see plenty of translators admitting things like, “I don’t know what a server really is, but my employer is happy with my interpretation about it.” From my observation, this seems to be typical and mainstream. In China’s case, the lack of dual majors or a culture of hybrid specialization makes this even more challenging.
In my own informal surveys of translators and translation students in China, I have found dozens of people who don’t know what a “corporation” is. If you are working for a US corporate counsel, you aren’t going to be able to communicate meaningfully if the “corporate” part of that is a mystery to you. This means that a typical business planning on relying on translators to interface in this way will find a total lack of talent. As a result, law firms and corporations tend to use professional translators for very low-level tasks such as personal emails or television subtitles.
Bucking this trend, a large “lawyer-linguist” movement has been emerging in the translation industry, focused on people who develop dual expertise in both law and language. Initially, this was adopted in Europe for bilingual legislation, reflecting a realization that translation work should provide more than bilingual text; it should provide value and usefulness. In the case of legislation, this would be fully functional laws that can be used equally in multiple languages, something that has traditionally been impossible.
Following the European lawyer-linguist movement, translators should also be able to provide clients with value by applying translation so that it is useful to the parties, not just in law but in a variety of essential business fields. As I argue above, this means that translation is about more than just the words, it’s about providing parties with real value.
