Translated on—but Not Necessarily by—a Computer

Would it surprise you to find out that today’s translators spend a lot of their time on a computer? This may not be the first image that comes to mind. Perhaps you imagine a person with a heap of dictionaries and reference books. At the most, they type up their carefully crafted translation before they send it off. Yet many translators use software tools to ensure their translations are consistent and their client’s formatting and code are preserved. As always, by translation, I will be referring to producing written content in another language, as opposed to interpreting oral or signed communication.

You Mean Machine Translation?

Of course, using computers to translate isn’t a new idea. Automatic, or machine, translation has been around since the 50s, and it’s made considerable progress recently. Does that mean that human translators simply use machine translation in their work or that machine translation is supplanting human translators? Not quite.

A lot of translation nowadays is commercial — that means business, scientific, medical, legal, patent, or other kinds of non-creative writing need to be made available in additional languages. For this kind of translation, consistent and accurate phrasing is paramount. To help with this, many—but not all—professional translators use special software that stores glossaries and previous translations and gives the human translator suggestions on terminology and phrasing. This software may be configured to include suggestions from automatic translation providers, but in many circumstances, they should not be used due to confidentiality requirements or the quality of machine translation.

Why Translators Have to Worry About Tags

Most texts nowadays are produced on a computer, but can we just enter the translation in a word processor? For some texts, that might work, although you would lose the benefits of consistency and accuracy that specialized translation software helps you with. However, sometimes what you need to translate is not a linear text per se. Other kinds of content, such as webpages, user interface strings, or desktop publishing layouts, also require translation. In that case, we should care to preserve not only the message but also such elements as:

• hyperlinks
• formatting
• non-translatable variables

That’s another reason translators may choose to use specialized software, which recognizes any XML or HTML tags your content might contain and helps preserve them in the appropriate spot in the translation.

In the end, translators who do not use automated, or machine, translation in their work often rely on specialized translation software to produce consistent content that reflects not only the meaning but also the form and functionality of the input text. Writing and content generation have changed in the last couple of decades—why shouldn’t translation?