Translation Techs VS Human Translators

From the Guardian

By Martin Williams

Does accurate translation always require a human eye?

Swansea’s councillors must be painfully aware of the dangers of technology in the translation industry. When a translator’s email reply landed in their inbox in 2008, the Welsh sentences were duly printed on a road sign. It read: “I am not in the office at the moment.”

But recent advances in technology are now helping to break down language barriers and revolutionise the role of traditional translators.

When Google Translate was launched in 2006, basing translations on hundreds of millions of online texts, it raised a crucial question for the industry: will technology take over? Now, as Microsoft prepares to unveil its Star Trek translator – a Skype service that promises to understand spoken words and translate them into another language, speaking them back in real time – that question seems more relevant than ever.

Pre-launch demonstrations of the app have been impressive, making only a handful of mistakes.

But the hype surrounding new technology does not mean computers have all the answers, according to Andy Way, associate professor of computing at Dublin City University. “You’re more likely to have everything else in Star Trek before you ever get a universal translator,” he says. Although enhanced technology is changing our approach to translation, the traditional translation industry is safe for now, he adds.

Experts divide translation technology in two distinct categories: machine translation (MT) which relies solely on software, and computer-assisted translation (CAT) which is simply used as an aid for translators. And although both are developing rapidly, translators say that only the CAT method produces high quality results.

Nataly Kelly, author of Found In Translation, explains: “Professional translators take great care to ensure that the message resonates with a foreign audience as the original author intended it to. Machines still lack the ability to do this. A machine doesn’t have a sense of humour, or the ability to choose the perfect words for a target audience.”

Professor Philipp Koehn, chair of the machine translation school of informatics at the University of Edinburgh agrees: “Automatic spoken translation is a particular problem because you’re working with two imperfect technologies tied together – speech recognition and translation.”

Koehn adds that there are worries that an over-reliance on MT apps could mean people rest their trust too heavily in a small handful of huge corporations.

“It’s certainly a concern for us in the research world, so we put a lot of effort into developing open-source software,” he says, citing his project Moses. “There’s currently not that much money being made with machine translation, but there are big efforts by Google and Microsoft and the potential is there.”

But globalisation has meant that, for some organisations, there is simply no other option other than to use new technology. Angelique Petrits, a language officer at the European Commission, says the organisation is responsible for translating 2 million pages into 24 different languages every year. “It wouldn’t be able to fulfil its mission without up-to-date translation technology,” she says.

“Technology is not meant to replace the human translator, but to speed up his work by automatically replacing strings of texts which have been already translated. In the best case scenario the human translator becomes an editor who is always responsible for the end product.

She adds: “Technology is a tool that helps dealing with scarce resources of translators, by speeding up their work and allowing them to concentrate on the essential. It also contributes to the consistency of terminology, crucial in EU texts.”

Like most professional translators, Lloyd Bingham, director of Capital Translations in Cardiff, uses CAT methods to make his work more efficient – but it does not replace traditional translation skills. “Translation technology is very much like Marmite to industry professionals,” he says. “The average freelance translator is less likely to hold it in such high regard as a large translation company – it’s designed to save translation companies money, while freelancers who work with it are expected to give discounts since such technology makes translation more efficient and productive.

“Generally speaking, translators enjoy working with translation technology provided that it makes the job more efficient and helps us to better meet our clients’ needs. Resources like Google Translate don’t do that. If you want the gist of a sentence, that’s fine. But never rely on it to translate a professional document, let alone one for publication.”

“CAT software saves segments for future use in what we call a translation memory,” he explains. “So if we come across a document for translation with similar wording in the future, we can use parts or even the whole translation of a certain sentence again. This makes translation a lot more efficient and crucially more consistent, as it means we are using the same terminology, style and register for the same client.”

Bingham adds: “Even though we are charging less because of the discounts, we are still working more efficiently and processing more work, so it’s swings and roundabouts. Plus, some people are coming round to the fact that Google Translate produces rubbish, so this is actually fuelling a demand for professional human translation.”

Technology, it seems, is having a deep impact on the translation industry, but is not about to eradicate it. Andy Way points out: “There is just so much translation to be done – people have estimated that only around 5% of what needs to be translated actually is – that good translators will never be out of a job.”

“What’s happening is their jobs are changing,” Koehn adds. “They are increasingly more about being a content editor than a translator.”

However, no matter how much the translation industry develops and improves, no translation will ever be as good as understanding the language for yourself, says Nataly Kelly. “When you learn words and phrases, you also learn cultural values. Many words do not have a direct translation, because the concepts simply do not exist in another culture. Language reflects society’s constant evolution, which is why it’s still easier for humans to keep up than it is for machines. “There is simply no replacement for learning a foreign language.”