Hardly a month goes by without another tech company promising to provide automated, flawless person-to-person “instant translation”. We have seen Skype Translator, NTT Docomo and a parade of over-hyped headsets – none of which actually did anything that impressive.
The latest in the line is Waverly Labs “Pilot earbuds”, which sounds remarkably like every other attempt to replace interpreters – listen to what is said, run it through an online Machine Translation engine, run the result through a speech synthesiser. Hey presto, you’re done!
In theory, it should work well. In fact, if interpreting was all about transferring words or ideas between languages, there would be no particular reason why computers could not, eventually, take over. Contrary to popular belief, computers are actually making great strides in understanding strips of language and matching them with common ways of expressing the same idea in another language.
But it takes all of ten minutes at a conference or doctor’s appointment or even in a library to see that language transfer is only one part of interpreting. Here are some common examples of what interpreters actually do.
- In hospital appointments, wherever cultural differences are causing a problem, interpreters spot the issue and communicate what clients are saying in a way that that is relevant and faithful but will keep things moving in the right direction.
- In mental health settings, interpreters offer professionals the information that they would instantly know if the patient spoke their language but can’t perceive when there is a language difference.
- In conferences, interpreters notice when speakers are being accidentally offensive, take into account what they were trying to achieve and create a version that does what they want, without the offense.
- In business negotiations, interpreters take into account that interpreting will lengthen proceedings and make people forget important information so they shape what they say to make the negotiations as effective as possible.
While it might appear that interpreting is a language job, with people attached; just a little experience teaches us that it is actually a people skill with language attached. Interpreters read the intentions of the speaker, manage interaction, take into account audience reaction, build rapport, explain cultural difference and more.
Yes, interpreters deal with language but they quickly realise that the language is almost always wrapped up in layer upon layer of social, cultural, and political complexity. Those are the layers that only a human can understand and, when it comes to the kinds of meetings where interpreters usually work, finding a route through those layers is the only way the event to work.
In truth, the kinds of work that will be taken over by computers are the kinds of work where we have traditionally relied on phrasebooks, dictionaries and, yes, online machine translation engines like google translate. We should all be glad to have computers and headsets to make it easier for us to ask for a beer, get directions to the nearest hospital and figure out what that tweet from your French-speaking colleague was about.
Frankly, using professional interpreters (or translators) in those situations would be a colossal waste of resources. Far better to use them in situations where getting it right will make the difference between life and death, negotiation and division, profit and loss. Machine translation is great and we are still testing its limits but, when it comes to bringing people together you can’t beat, well, people. And if you want people who know how to cut through the jungle of cultural and social differences, you want interpreters.