BY: CAROLINE ROLF
Travelling overseas with my family used to mean packing my pocket dictionary and studying phrases on the airplane (after looking up curse words, of course). After flipping back and forth through the pages and receiving a little guidance from my father on pronunciation, I could almost construct a proper sentence. Today as I continue to travel and meet new people abroad, communicating with them is as fast and easy as typing a phrase into Google translate or asking my smartphone Siri.
Sound like a familiar problem? Well the team at Waverly Labs has created a device that will be able to help. Come fall, Pilot, an in-ear translating device is expected to hit the market, allowing two people to communicate speaking two different languages while still clearly understanding each other. Real-time computer translation could be one of the most radical technological achievements, breaking down language barriers, accelerating interconnectedness and opening doors to new cultures. Without a doubt, portable machine translation is faster and far more effective than a dictionary, but what it still lacks is accuracy, meaning and delivery.
Photo: Waverly Labs
The thing about real-time translation, as ground-breaking as it may be, is that translation is only part of the issue. Translation will allow for comprehension of content. For simple sentences and words between grammatically similar languages, these work pretty well, but try to translate a complex sentence from Chinese to English, for example, and the results are basically gibberish.
In this Wall Street Journal article, author Alec Ross argues that over the next ten years, machine translation will enable dialogue between dozens of languages, eliminating barriers. Surely this is true for simple phrases like “Where can I get coffee?” but the real-time flawless translation that Ross describes is likely a lot longer than ten years away from happening.
For starters, language is vastly different from culture, but is heavily influenced by it. Without realizing it, we reference aspects of our culture in our everyday speech, some of which can’t be understood by foreigners. If I mentioned that I live in “The six,” for example, an outsider understanding that would depend on their familiarity with rapper Drake’s reference to the city of Toronto. The reference as opposed to the literal meaning could be programed into a machine, but language moves quickly and constantly updating cultural references would be tricky.
Even a seemingly universal phrase can differ from culture to culture. This article from Tech in Asia highlights that in an American office, the sentence “I’ll think about it” means the speaker will consider it, however in a Chinese office, the same phrase means “No.” A human translator would be able to explain this with knowledge of the cultural background, while technology would have a difficult time distinguishing between the literal meaning and something else.
The meaning of our words can also vary based on the context that they are used in. “Thanks a lot” is an example that could mean you are gracious for a friend’s gift or quite the opposite if you’re blaming your friend in a sarcastic way. Human interpretation picks up on the speaker’s tone and context, whereas this information is likely lost in a machine translation.
A device like the Pilot translator won’t be able to break the language barrier—at least not in the near future. Languages are structured differently, so a machine would need to hear a full sentence in order to translate properly, meaning it can’t work properly in real time. It will never be perfect, but it will give the idea of perfection and that’s where the danger lies. We trust the software to translate our conversations correctly, but as we have seen in media blunders, the consequences can be catastrophic. It can even be difficult for a human translator to rephrase things like humour and accents, and that’s just for casual conversation. True communication includes context, human contact and empathy, something a machine can’t understand accurately. There was a time when every method of communication seemed nearly impossible or very far off, including the printing press and the cellphone. This is to say that real-time language translation could be a real possibility in the future, but maybe we should stop overestimating computers and underestimating the human brain.
