Suggestions for successful Japanese in-country review

Bell: I don’t know much about Japanese culture. I feign to say I know nothing, though I once read Memoirs of a Geisha, but I don’t really think that counts. It’s a shame because from everything I’ve heard, Japan is a great place — steeped in tradition, elegance and art. What I do know about, though, is translation. I know how to make sure a document’s prepped for the best translation possible, I know how to select the right team of translators to do it, and I know how to manage quality control and assurance so that the translation reads as well as it does in the original. The processes involved are relatively foolproof and tend to be adaptable across languages.

When it comes to Japanese, though, something — anything, everything — almost always goes wrong. And it’s not just me. Most of my colleagues will tell you that translation into Japanese is hands down the most difficult work we do. In the world of Japanese translation, no news is not good news. Feedback is often vague. And in-country review (ICR)? Well, let’s face it,  ICR of a Japanese translation is a royal pain in the oshiri.

All multilanguage vendors (MLV) have gone through it. A painstakingly-prepared translation is delivered to your client, who then sends it off for ICR. Ideally, the reviewer speaks English in order to compare the translation to the original, but sometimes he  or she doesn’t. Weeks pass, even months, and then it comes — sometimes long after the bill has already been paid and the files have been moved off the hard drive. The in-country reviewer isn’t pleased. “The translation is bad,” he or she says, and oftentimes says no more than that.  It’s frustrating. You write for details — was it word choice, was it grammar, did you accidentally translate into Korean? Nothing comes. Meanwhile, you sit in your office, panicked that your company’s reputation will crash around you, all because some person in Japan doesn’t know how to elaborate.

Of course, this is how you see it if you know nothing about Japan. But if you do know something about Japan, you see how the same culture that gives us geishas, calligraphy and plum wine also contributes to poor ICR — or what we call poor ICR in our Anglo-Saxon construct, that is. When we start looking at ICR as a cultural process and not a linguistic one, feedback — or lack thereof — starts to make more sense. The Japanese are, of course, Japanese. We can’t expect them to act like anything other than themselves. In our American culture, we are so delineated, so black and white. We are demanding, confrontational, young. Our culture is a child compared to theirs, and we shouldn’t be surprised if sometimes the Japanese don’t feel like they shouldn’t simply pat us on the head and sigh. It makes sense that we would want everything immediately detailed out and that they would be a bit more patient and reserved. If we consider culture to be an integral part of translation, then we must consider it to be integral to proper ICR as well.

Enter Maureen McCarthy, an employee whom I’ve asked to cowrite this article. Maureen also happens to be In Every Language’s Japanese expert. Call her the Encyclopaedia Britannica to my Memoirs of a Geisha. Everything I don’t know about Japan, she does, which is reason #412 why we like to have her around the office. Maureen, am I crazy? Could the delay in ICR response time have anything to do with the Japanese not wanting to rush things, with their being less confrontational than Americans are?

McCarthy: In my experience, the delay time has more to do with the hierarchy that exists in a Japanese workplace and the need for important decisions and documents to go through a lot of people. For example, when I lived in Akita, a friend of mine was planning to go on a business trip to Tokyo with some other colleagues. The trip was an annual affair with many people involved, and for months my friend had been eagerly awaiting this retreat. She had cleared her involvement with her immediate boss and with the head of the office. However, the trip was to take place in May. In Japan, there are often big employee shifts in the workplace every April, as this is when the academic and fiscal years begin in Japan. In this instance, the trip coordinator moved to a different branch and a new coordinator came on board. This meant that many aspects of the trip changed, including my friend’s involvement. The new coordinator thought she was not a necessary member of the group, but instead that she should remain at the office as she had many responsibilities there. Unfortunately, she was not informed of this decision until two weeks before the trip. This is not through any fault of the new coordinator, nor hers or her boss’s. It is simply that major decisions often need to go through a lot of channels in the Japanese workplace. In this case, the decision was especially difficult because key players in the decision-making process changed, and the situation had to be reevaluated from the beginning.

More specifically, with ICR response times, there could be a number of factors at play, such as the people involved with the decision-making,  the steps it goes through on the Japanese end, and even the time of year — March, April and May were always very hectic at my previous workplace. It is even possible that the original reviewer thought the translation was fine, so he or she did not see it as a priority for his or her superiors to view. But when the bosses did eventually see the translation, they may have had a very different idea concerning its suitability and told the initial reviewer to write back saying they were not pleased. By now, of course, weeks or maybe even months  have passed before you receive the e-mail saying the translation was not up to the standard they expected.

Bell: So could that contribute to the fact that the reviews are often incomplete by American standards? ICRs are just a waste of time if they’re not helpful. Instead of a simple pass/fail — as though translation were a ninth-grade gym class — the reviewer should score the translation on, say, a scale of 1 to 5, rating clearly specified factors such as grammar, spelling and non-subjective word choice — think translating neko as dog instead of cat. These factors should be agreed upon between the client and the reviewer ahead of time, and if subjective factors start to come into play, these factors should be presented to the language service provider (LSP) so the LSP can better understand the client’s needs. 

ICR itself has one of two goals. Some companies use ICR to help select a translation vendor. Multiple LSPs translate a sample and then an in-country reviewer decides whose translation is the best. In these instances, it’s essential that the reviewer speak the original language so he or she can tell if any “errors” are issues with the translation or problems that were also in the original.

 The other common goal that ICR sets out to accomplish is to make sure that the translation is ready for its target market. In this way, reviewers are the governor on the go-cart, the childproof cap on a prescription, the airline agent who scans your ticket before boarding. In other words, the buck stops with them. That’s why you shouldn’t pick just anyone to perform your ICR. If reviewers are your final layover on the way to happy translation land, make sure they know what they are doing. In both instances, it’s essential that the reviewer be familiar with the content matter and, of course, the target language, country and culture.

 That’s why it’s so disparaging, as an LSP owner, when we receive back the occasional, negative ICR. All negative ICRs are brought immediately to my attention to make sure the issues found are addressed. But if no specifics are provided, well, to quote George Costanza from Seinfeld, “I’ve got nothing.”

I appreciate what you say, Maureen, about how it may be the reviewer’s boss and not the actual reviewer coming back with ICR comments, and I think that’s part of why Japanese ICRs can be so non-descript — because the person reporting on the review might not actually be the person with final say. But are there any other cultural factors that keep the details on the kibosh? Why are Japanese ICRs so ambiguous? And, more importantly, is there anything LSPs can do to help a reviewer provide the detail we need?


Formality and culture

McCarthy: The Japanese are often described as extremely polite and formal. While this is true to a degree, there is also a strong inclination among many to be nonconfrontational. If a conflict can be avoided, it usually will be. This is, of course, wonderful in many ways, as it means less fighting and troubles over small things, but it also has a downside in that we in the West often expect to talk through any problems. We Americans usually count on others, especially in a working environment, to tell us if they are displeased so that we can work through it.

When I first began learning Japanese, all of my Japanese friends and colleagues were very supportive and encouraging. They would always compliment me on my language ability and never corrected me or told me when my Japanese was incorrect. For me this was wonderful, and everyone’s kindness and encouragement really motivated me to do better. On the downside, when I began to study Japanese formally, I discovered I had been making mistakes all along. No one wanted to be the bearer of bad news or make me feel bad by correcting me.

When an ICR is undetailed, this most likely means that no one wants to be the bearer of bad news. When a reviewer feels that a translation is unacceptable, he or she probably does not want to dwell on the details as that may be uncomfortable and potentially embarrassing to one or both parties. This could explain the very brief answers and the lack of a proper, detailed review. I am not sure if there is an easy solution to this problem, as so much of it stems from differences in our cultures, but I think, Terena, your suggestion of rankings on a scale, or a list of questions, would be a good approach to take. If the reviewer had a survey or form to complete, asking specific questions and having a list of choices, he or she may be more likely to respond and to give LSPs the details they desire.

Bell: Wonderful. It makes perfect sense that the reviewer wouldn’t want to be overly harsh or hurt anyone’s feelings with a review. I think we all agree that there’s a difference between constructive criticism and just being mean. But in translation, a mistake isn’t always a mistake. At home, I may sit on a sofa in my living room while you sit on a couch in your den. Language is often subjective. I hate to think that every time an ICR doesn’t go well, it means that the translation was poor. Sometimes it simply means the translation wasn’t perfect, with the meaning of perfection taking in a lot of subjective elements. Reading your response, though, Maureen, I think that no news, or detail, at least in this case, is bad news — that a lack of detail in the ICR means that the errors were all non-subjective, something made by an early language learner, like you said. Is that correct, or am I simply misunderstanding? Is it always a matter of bad or good, or could lack of detail in ICR mean there were subjective differences instead?

McCarthy: It doesn’t necessary mean that the errors are the translator’s fault, although that is a possibility. It is likely instead that the language the translator used could simply be different than what the reviewer wanted or expected. There are many different levels of formality in Japanese. The way you speak to your boss is not the same as the way you speak to a colleague, which again is different than the way you would speak to your family members. If a document is being translated for a company, it is important to clarify what level of formality is to be used. For example, if the document is something that’s given to customers, most likely the customer will be addressed in very formal Japanese. If, however, the document is of a more familiar nature, it need not be as formal. It is also important to remember that Japanese is often not as direct as English or other Western languages. Take, for instance, leaving work for the day. In English we may say something along the lines of “See you tomorrow!” In polite Japanese, however, one would say osaki ni shitsure shimasu (excuse me for the rudeness of going before you). If you’re not leaving work, though, and instead leaving a gathering of friends, it would change. While the English is still “see you tomorrow,” the Japanese would change to mata ne (see you later), ja ne (see ya), or maybe even mata ashita (see you again tomorrow). Although situational differences are important when translating between all languages, Japanese has an abundance of them. Not only is Japanese often more florid than other languages, it also has a daunting amount of set phrases that are used in specific situations, such as the aforementioned leaving work. For ICR this may not mean that the translation is incorrect, but that there may have been a different phrase that the client wanted to be used. It could also be that the translator put the document perfectly into Japanese, but that the original document was very brusque and to the point, whereas in Japanese it would normally be written in much more elegant terms. This would not really be a translator error, but merely a lack of communication in terms of what the client really wants, as well as evidence of the great difficulty of translating between two such different languages. Perhaps, then, the only true solution would be more in-depth collaboration with the client, such as asking a series of questions before beginning the project or touching base with the client periodically throughout the process.

Translating vs localizing

Bell: Another solution I can think of is to make sure your client wants translation instead of localization. Now, I’ve heard the argument that a really good translation is localization, but when clients entrust us with their documents, we as translators must take care not to overstep our bounds. It’s very easy to say that if a document is to the point in English that we need to soften it up in Japanese. But when a client asks for translation, as opposed to localization, we also have more of an obligation to be true to the original. There may be a reason why the text is more direct. Perhaps the author wants to unquestionably drive home his point, or maybe he or she is upset and wants that emotion conveyed through the text. In those cases, completely rewriting the text to softening the tone and formalize the language would be inappropriate.

This is why it’s so important to get as much information as you can from your client before you begin. If the tone of an original document could be construed as rude in Japanese — or even simply neutral — we should point this out and find out how much of it is — and isn’t — intentional. At this stage, we’re doing more than simply translating; we’re making the document local. The difference between translation and localization becomes a hazy gray zone, and it takes cooperation between the translation team and the client to find out where along that spectrum the translation should hit. ICR should definitely be involved in that process; if the reviewer doesn’t know what the translator is striving for, then he or she won’t know how to judge the outcome.

Good translation cannot be created in a vacuum. In this regard, Japanese really isn’t all that different. In fact, because the Japanese culture focuses more on the team than the individual, a collaborative approach to translation should be easier for Japanese than for other languages. The processes applied to those other languages still work here: Get as much information as you can up front. Understand not just the words, but the intent, meaning, and goal of a document. Take culture into consideration.