What Would a World Without Language Barriers Look Like?
My dad has a story he likes to tell about one of his friends, a scientist. The scientist was giving a lecture in Japan, and opened with a joke that lasted a couple minutes. After delivering the joke in English, he waited for his translator to relay it to the audience. The translator spoke for only a few seconds, and then the crowd burst out laughing.
After the presentation was over, the scientist asked the translator how she managed to distill the humor of his joke down into such a concise form. She shrugged and said, “I said that the American visitor just told a very funny joke, and that they should all laugh now.”
The scientist’s story illustrates the subjective, human quality of translation. Moving between languages is rarely a matter of transposing literal meaning; it requires the constant triage of unexpected inputs, endless judgment calls, and some social awareness. In other words, it’s something that humans are cut out for, and that computers are not.
With enough coaching, could a computer become as good as a human? At Microsoft, hundreds of humans are trying to train a machine to listen, translate, and then speak. Last December, the company announced the limited release of Skype Translator, which can translate a conversation between two people videochatting in different languages, in real time. The software, which is still invite-only, can handle English, Spanish, Italian, and Mandarin. Google, too, has a smartphone app—free and available to the public—that can transcribe spoken text in one language, translate it, and then speak the result aloud in another. It’s not hard to imagine Google embedding this technology into its own videochat platform to a similar end.
Earlier this week, I had the good fortune to test out Microsoft’s new software. The company paired me with Ignacio Horcada, a translator who, as Skype Translator had it, had previously worked “in a context of the species specialized in nuggets”—or, as humans would say, he used to translate web pages. Now, he helps correct Skype’s translation errors.
I went into the call expecting a normal conversation, but soon realized that it was an entirely different mode of communication with its own set of constraints: Once I started speaking, the software waited until I stopped to transcribe my words in a sidebar and then speak them to Ignacio in a computerized voice. This meant that we couldn’t interrupt each other. It also meant that thoughts were best expressed in one- or two-sentence bursts, not paragraph-long rambles—a lesson Ignacio had clearly picked up from using the software so frequently, and that I was slower to catch on to.
Small annoyances piled up in the form of mangled meanings and misheard words, but I hung up on the call with a sense of Skype Translator’s promise. In light of its amazing premise—we were speaking different languages but making sense to each other!—those mishaps were easy to forgive. (After all, humans make mistakes too: A few weeks ago, Cleveland Cavaliers center Timofey Mozgov answered an English-speaking reporter’s question with 15 seconds of rapid-fire Russian, and didn’t realize his mistake until after the reporter’s intrepid follow-up.) Skype Translator will only get better; it depends on machine learning, a process that evaluates its own outputs and makes adjustments accordingly. It’s what has enabled mapping apps and Google searches to improve as more people use them, and the same will likely happen to live translation.
For now, Microsoft’s focus is on how this technology would fit into the lives of consumers. On its website, in its promotional videos, and at live events, the company has been dwelling on what the technology offers average people: Students can talk to peers in classrooms around the world and travelers can sync up with locals before trips abroad. It’s not hard to imagine other everyday functions—as a journalist, technology like this drastically widens my pool of potential sources—but one gets the sense that there’s a bigger push into the business world being planned.
“Of course, the day our CEO showed it on stage, we’ve been flooded with a lot of interest from many of our enterprise partners asking about the business implications of this technology,” says Vikram Dendi, the strategy director at Microsoft Research. Dendi has been helping oversee the development of Skype Translator, and is not hesitant to note its current limits. “I would want to make sure that if we develop into a more mission-critical type of scenario, there is that level of capability and maturity [first],” he says.
Distributing instant-translation software to Skype’s 300 million monthly connected users makes sense not just because lay people have more tolerance than executives for imperfect translations, but also because it’s a way of putting it (as well as videochatting itself) on their radars for potential use in business. “As the habit develops, they will ask their organizations to empower them to do the same, or even better when needed with high-end video conferencing systems. It will increasingly be a bottom-up push as people adopt it in their private life,” says Nicolas de Benoist, a design researcher at the office-furniture manufacturer Steelcase.
Skype Translator, if perfected, would likely be of use to large multinational corporations that depend on the presence of accurate translators for sensitive transactions. But perhaps even more potential lies in applying it to developing economies—places where businesses are smaller, people can’t afford professional translators, and, most important, English hasn’t taken hold as the default language of commerce.
“As markets expand into areas where English [is less common], in other words newly developing markets like Africa, parts of Asia, and critically in Latin America … I think this will have a really profound effect,” says Dean Foster, the founder and president of DFA Intercultural Global Solutions, a consultancy that educates businesspeople about foreign cultures. For example, even in economically mature China, 400 million of the country’s 1.3 billion people had learned English, as of 2006, the most recent year with dependable data. But of those 400 million, more than half said they “seldom” spoke it and only between 5 and 15 percent said they spoke it “often”—which leaves a lot of people who are likely more comfortable doing international business in their native tongue.
As useful as this technology will be to businesses both small and massive, though, it has yet to translate one complex component of global interaction: culture. “Even if they got [the language] 90 percent right, they’re not dealing with the issue of culture … For us, what we call cultural fluency as opposed to language fluency is probably the more critical determinant of business success,” says Foster.
Foster might be biased, as the head of an intercultural consultancy with a degree in sociology, but stressing international understanding is not just a touchy-feely sticking point. Missing cultural cues can lead to material losses: In Brazil, for example, touching the thumb to the index finger—the American sign for “OK”—is a profane gesture. “There are stories when American businesspeople go down there after negotiating a deal for months—a major, major business deal—and then they sign the contract and flash the ‘OK’ sign, and they’ve just insulted the CEO of the Brazilian company in public with cameras flashing and everything on the media the next day,” says Foster.
The reason real-time translation might get blindsided by cross-cultural hiccups is that it can give the illusion of understanding. “It’s easy to get your head around, ‘If I don’t speak the language, I’m obviously not communicating.’ It’s harder to get your head around the fact that, ‘If I don’t understand the culture I’m not communicating,’” says Foster. As an example, he points to Americans and Britons, who expect to get along in business scenarios because of their shared language, but often exhibit very different management cultures. (Plus, England has its own offensive hand gesture that Americans might unwittingly throw out.)
And one thing that might be lost in a world where translations become sufficiently smooth is the conscientiousness (or perhaps flattery) involved with learning a client’s native tongue. Yui Kong Heung, who worked as a translator in Hong Kong for a major Japanese electronics manufacturer, found that he was more likely to win over his bosses when they knew he spoke their native tongue deftly. “The Japanese are very open-minded and easygoing once they know that you speak the language,” he says.
As Skype’s instant-translation software gets fine-tuned in the coming years, bigger questions might actually be asked of its medium, videochatting. Chris Congdon, a colleague of Nicolas de Benoist at Steelcase, told me that while the idea of videoconferencing polls highly among businesses, its adoption rate has been lower than expected. “What I’m interested in with these technologies is whether they can get to a place where they become so transparent that it doesn’t feel like there’s a layer of something happening between you and the other person who you’re trying to build a relationship with,” she says.
There are other obstacles to videochat’s adoption as well, says de Benoist. Cohesiveness arises when people are sharing a space, from the layout of the room to what the weather’s like outside. That’s hard to replicate with video. Also, meetings themselves might not be when some of the most important information is exchanged. “It’s actually when people leave the room … that they’re sharing informal content, and the people who are more introverted are saying, ‘I had this that I wanted to say, but I didn’t feel like it because it was too formal for me,’” de Benoist says.
Real-time translation in videochat might seem as distantly futuristic as some of the other ideas I heard about the future of business meetings—Congdon spoke of cameras that capture participants’ body and gestures, and Dendi mentioned a possible down-the-road integration with HoloLens, Microsoft’s augmented reality goggles. And it’s safe to say real-time translation is far off. But then again, there’s a time when every mode of communication seems far off—Slack, email, and the telephone included.
Posted, but not written by, Louis Sheehan