How to avoid context-related translation errors

The value of high-quality translations has been proven time and time again. Yet there’s a crucial issue with their execution, that, even in our age of almost almighty technology, has yet to be resolved.

The issue: context—or rather, the lack of context that’s being provided to translators.

There are enormous benefits to be gained by providing your linguistic team with style guides, glossaries, screenshots—as many reference materials as possible. But even with all these materials provided, the amount of context available is limited by the translation process itself.

As it stands, translations are carried out within a translation platform—not within your individual document/app/platform. This means that the translator does not see the source text in context. They’re instead presented with a list of detached phrases, as in the screenshot below:

This format deprives translators of any context. They’re faced with a list that may contain both one-word strings (“Reset”, “Copy”), and larger blurbs (such as privacy policy text), combined into one huge—and confusing—segment.

Of course, linguists have a general understanding of your product. But even if they’ve had an introductory look at the copy, they cannot know exactly where each of the strings they’re translating appear in the UI.

In one scenario, linguists incorrectly translated the string “home” as “house.” The website being translated belonged to a furniture retailer, but, unbeknownst to the translators, the string in question was actually a home button.

So, how can we avoid errors resulting from a lack of context?

  • Provide translators a screenshot or developer notes for each and every string (the translation platform Transifex allows you to do so)
  • Perform a professional linguistic testing of your product/app after the translation has been implemented

So far, on-device linguistic QA is the only bulletproof solution for companies striving to release a seamlessly localized product to their target markets. Due to the context limitations faced by translators, linguistic issues—as well as layout/display issues—can be identified only during the QA step. This step involves a QA tester navigating through the test plan and auditing every possible screen, in a manner mimicking how a user would interact with the product. This is the best time to catch any inconsistencies or mistranslations (as in our “Home” example), since the translation is presented in context.

At e2f, we offer thorough linguistic testing to ensure that your product (software, website, app—you name it) is in it’s best possible form for an international audience. For more details on e2f’s services, check our Linguistic testing page or get in touch at [email protected].