Beauty is in the Eye of the Beholder: Subjectivity in Translation

So many in our industry are in love with language. Multiple degrees in linguistics, advanced studies, even language-related hobbies are common for many professionals in the language services industry. Lively conversations where people compare various languages, dialects, characters are common – it’s not work for them; it’s a passion. How lucky we are to be in an industry where so many people care so deeply about the service they provide.

In my last blog post, I discussed the challenge of trying to plan perfection, which, by its-nature is in conflict with the requirements of a continuous-flow process. Some of this disconnect is driven by the very passion for language that is pervasive in our industry. I remember a conversation I had with Don DePalma of CSA in the mid-90’s sharing my frustration about the lack of common quality standards and the inconsistent if not capricious nature of reviewer feedback.

I wish I could report better progress since then. Certainly we have made steps in terms of glossary automation, consistent use of style guides, and many other great tools professional linguists can use. Yet at the end of the day, when a client reviewer has strong preferences, this becomes the de facto standard of “right” and anything else is “wrong” or “bad.”

At a recent Localization World presentation, a manufacturing company shared their methodology of QA guidelines provided to reviewers that clearly define categories and severity of errors. Combined with a consistent scoring methodology, this moved the review process from the purely subjective into something that looks to be reasonable, repeatable, and scalable. But this global company is able to enforce these standards and processes on their employees around the world. The reality is that this company and this methodology represent a small slice of the global demand for translation. 

We often ask clients, “Please connect us with your reviewers so we can set guidelines and establish clear expectations.” We are often told that reviewers are not available, or at least none that will be consistent from release to release. Yet the feedback comes, changing approved glossary terms, not following the style guide, or making inconsistent, preferential changes and varying from reviewer to reviewer. Often changes are made because the reviewer does not like or is “correcting” the source. Aside from the natural frustration, this affects the translation memory so the client loses quality, consistency and cost benefits. Everyone loses.

Okay, it’s a bit of a whine. This is a service business and we our proud to take care of our customers – especially in difficult circumstances. The reason for my rant is that we have a business to run, and these types of changes affect the bottom line. Sure, we can charge more – then invariably someone points out that a different company is far less expensive. This is the heart of the matter: in a business where the output is variable and without a clear standard of measure, customers primarily compare by price. We are the victims of our own inability to clearly define and articulate language quality standards and gain broad acceptance. As a proxy, customers rely on the brand of the translation company, or sometimes on trusted personal relationships, to provide the confidence needed to engage services. Or they use translation tests, a subject for a few words another time.

For example, we know that the minimum standard of competence for an accountant is to be a Certified Public Accountant (CPA) in the US or a Chartered Accountant in the UK. Attorneys must become members of their bar association. This is not to say there are not valuable translator associations. Membership certainly implies that the translator regards his or her profession with commitment and has competence but there are usually no or little barriers to entry. However, with rare exception, efforts to create standards neglect perhaps the most critical component of all: specialization. You wouldn’t knowingly see a podiatrist about a problem with your eyes or ask a patent attorney to defend your traffic violation. Yet many are willing to select a translator that has never encountered the term “flying buttress” or “equity derivative” and wonder why their in-country reviewer complains about the translation.

The beauty and complexity of translation is that it is ultimately a human endeavor. Humans are both variable and imperfect. Choosing the right humans to translate for you is perhaps the most important variable to nail down. Beyond that, we’ve established processes (two-step translation+editing) that are reinforced by workflow automation to reduce the instances of “bad” translation getting to the customer. We follow industry-standard best practices. We developed a statistically-driven QA methodology to constantly evaluate practitioners to keep everyone performing at peak levels. Our collective challenge is to educate our customers on the importance of specialization and how to measure quality in non-subjective terms. This will help us to improve highly variable project margins and, more importantly, continually elevate the level and consistency of the service we deliver. Everyone wins.

What do you think? Post a comment – I’d enjoy sharing a few words with you.