What's your most desired sci-fi gizmo?
Frequent flyers will probably go for beam-me-down technology or that elusive flying car, but universal language translation would probably be pretty high up most people's lists too. And unlike food replicators and warp speed, it looks like you might actually be able to see this last one in the real world sometime soon. (The odds of finally getting your flying car are better than you think too.)
In the movies, instant translation is accomplished by cute droids and all-knowing ships' computers, but here on present-day planet Earth, we're going to have to settle for a nice pair of ear buds.
Following years of machine translation breakthroughs by the likes of Google and Skype, a few months back, a company called Waverly Labs created a media firestorm by promising the imminent arrival of a $299 device that provides instant, in-ear translation for English, French, Spanish and Italian (with additional languages available for download). Here's the promo video:
So if you often travel to Paris or Milan, should you shell out the cash? At the time, a few commentators expressed skepticism about the specific technology and company, but the the most thought-provoking responses of all came from those concerned not with business or tech, but with linguistics. And their answer is straightforward -- the value of this sort of device depends very much on what you hope to do with it.
Why language is different
Computers can win Jeopardy these days. Surely, they should be able to tear down the language barrier too, right? But as University of South Florida language professor and translator David Arbesú points out on The Conversation, despite humans' near miraculous facility for it, language is actually harder than game shows -- way harder.
Given enough data and processing power, computers can and will get better at swapping out a word in one language for a word or phrase in another, he writes. But if you think that's what translation is, you're sorely mistaken.
"Replacing a word with its equivalent in the target language is actually the 'easy part' of a translator's job," he explains. "Translation doesn't - or shouldn't - involve simply translating words, sentences or paragraphs. Rather, it's about translating meaning. And in order to infer meaning from a specific utterance, humans have to interpret a multitude of elements at the same time."
Like what? The tone and volume of your voice, the situation, cultural cues, metaphors, and jokes, just to name a few. And that's not even mentioning grammatical complexities like rendering formal and informal forms of address (like tú and usted in Spanish) into another language.
What real-time machine translation can and cannot do
Without being able to parse these variables, the usefulness of currently imaginable translation tech is limited. It's probably not good enough, for instance, to allow you to fall in love with a good-looking foreigner while chatting over a cappuccino, a situation The Atlantic's Spencer Metsel imagines in detail in his awesomely titled article, "Could Two People Use Real-Time Translation to Fall in Love?"
It's a fun read with a great conceit that manages to painlessly slip readers a hefty helping of potentially dry linguistic detail, so check it out if you're interested in the topic. But Metsel's essential takeaway is the same as Arbesú's: current real-time translation is undoubtedly useful for finding a cafe or ordering lunch -- and worth considering if that's what you're looking for -- but that nuanced protocol droid of your dreams, the one who actually allows you to have a real conversation with a stranger, will still be science fiction for a long time coming.
Would you consider buying instant translation ear buds?