Quantcast
Membership Signup
Singularity University

Will It Matter If We Speak Different Languages In The Future?

The Rosetta Stone, an ancient Egyptian decree written in hieroglyphics, script, and Ancient Greek, made translation of hieroglyphics possible.

The notion of ‘foreign’ languages is beginning to disappear, thanks to ongoing developments by the likes of Google and Facebook that help the thousands of languages across the world be less foreign and more accessible. How? By making language translation easier than ever. In light of the technological progress in translation software made over the last decade, we may very well be on the cusp of witnessing the majority of language barriers between people groups disappear.

Not only does that mean everyone will be able to speak their natural tongue, learning another language will be purely out of interest or linguistic study, not a degree requirement for students or mandatory continuing education for international business people.

Leading the way in these innovations is Google, and as the search engine has become popular so has its translation service. Google Translate now has 200 million monthly users accessing the service through web pages, YouTube, and mobile apps. According to Google’s CEO Larry Page, Google Translate now works with 64 languages instantly and for free.

Not content with software that’s good but not perfect, the company has recently made a slew of improvements to Translate. Google uses a statistical approach in its translation service, which means it pulls translations from web pages it has indexed to find the best translation. But since last May, the company has allowed users to help improve results by editing translations. On its blog, Google also announced that Gmail now allows automatic translation of incoming emails, something that will certainly make both its personal and business customers happy. Finally, Google Translate was beefed up even more by incorporating example sentences from web sites that use the translated word or phrases, thereby providing greater context, just like some dictionaries include.

All of these developments are helpful for the printed word, but what about speech? Fortunately, you can get translations with Google Voice Search in Chrome simply by saying “Translate to”, naming the language, and then saying the phrase. The translation is in written form, but surely it won’t be long before Search will speak back to you. That’s because the Google Translate app does just that in Conversation Mode, allowing users to type or speak in their native tongue and then see or hear the translation in seconds.

Here’s a 2010 video explaining the backend of Google Translate:

Google is not the only one employing translation software. Last October, Facebook introduced a translation tool allowing users to instantly translate posts in foreign languages. Because the social network has such an intersection of people from around the world (900 million as of March 2012), translation is a high priority, which is likely why the company turned to Microsoft Bing rather than develop its own translator. Bing Translator recently replaced the once popular Babel Fish translator from Yahoo!. The Microsoft service currently translates 38 languages in Office, Windows Phone, and Kindle as well as Facebook.

But do people really want technology to do all the work for them? Perhaps a strong interest in learning a new language persists even in the age of the web?

To address that question, there’s no better place to look than Rosetta Stone. Founded in 1992 as Fairfield Language Technologies to develop translation software, the company grew to provide language learning packages in 30 languages. In 2009, Rosetta Stone became publicly traded alongside a marketing blitzkrieg that was quite successful…from an advertising point of view. A survey showed that nearly 80 percent of Americans are familiar with the brand, yet the company reported a 73 percent drop in annual sales by 2011 likely due to a handful of reasons, such as the general decline of software sales and competition, but it appears that that biggest reason is lack of interest in average consumers.

Even as the company wrestles with declining interest, Rosetta Stone continues to market its software as a convenient way to simulate immersion through video in order to learn a language. Now, most of us don’t think about language before we speak — we just talk. Yet, one of the biggest struggles with learning a second language is training your brain to think in that language. Achieving this takes years of practice and to be truly bilingual almost always involves immersion, usually in a foreign country. Otherwise, you are forced to continually perform the mental task of translating from your native tongue.

And that’s the point. As with many other cases, these kinds of mental processes will be faster and more accurate when it is done by computers.

Think of it this way: mathematics itself is a language and many a math student has struggled with its translation. But calculators have changed the way that people do math, and the number of functions that the average calculator can perform has grown immensely. Sure, it would be great if people had a better appreciation for the beauty of mathematics, it’s history, development, and all the little nuances and tricks that make it so fascinating (just as with foreign languages). But the majority of people need to use mathematics for very practical purposes, like balancing their checkbooks, so learning about Fibonacci sequences in school tends to muddle all the pragmatic math that they will actually need. In most cases, students benefit from learning all the details of math, but what they must absolutely know is how to use a calculator.

So where is translation headed? Technology is clearly pushing toward a universal translator — in other words, a language calculator. A universal translator will take all the mental calculations out of translation so you can focus on what you want to say, not how to say it. After all, you don’t need to recall from language class the history of the Spanish Empire, what the components of paella are, or the popularity of Flamenco dancing when you are in Tijuana and just need to find a restroom among the seedy night clubs (yes, I am speaking from personal experience).

And technological progress won’t stop there. At some point, we’ll want other translators and the possibility of interspecies communication will likely become of great interest. In fact, it’s already begun with groups like the Wild Dolphin Project and the Sarasota Dolphin Research Program aiming to do this by learning to decode dolphin communication. It may seem that speaking with animals is just a child’s wish, but we’re in an era when technology is making childhood dreams come true.

This year’s April Fool’s Day video from Google pokes fun at the idea, but you can bet someone at Google Labs has seriously considered whether it could work:

But in the here and now, what does all of this mean? Your best bet is to focus on one language, your native tongue, and master it in all forms. In the age of the Internet, communication is more important than ever, because more people are communicating and they are using a variety of means to do it (blogs, podcasts, and YouTube videos). On top of that, the web is persistent, so what you say can linger for a long time.

So when it comes to foreign languages, let technology take care of bridging the gaps between different tongues, dialects, accents, and slang. There’s nothing wrong with learning a language for fun, but the days of believing that you have to know a second language are over.

Reply

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

8 comments

  • why06 says:

    I agree and disagree.

    Will language software eventually master translating languages? yes.

    Does that mean a device will help you understand everything someone speaking another language is saying? No.

    For instance the sentence I just wrote: “someone speaking another language” that may be exist only as one word in another language. Lets say the word is poohbear. Well in this case it would be better just to learn the word instead of expanding the definition in mid sentence. This slows down conversation.

    To generalize I would say figures of speak, slang, and words should in some cases NOT be translated, but learned through communication. by doing this what one creates is an intermediary language, where both sides can understand some and infer the rest.

    Some puns, riddles, and things will only make sense in the native language. So I would say even with a perfect translator as always… information will be lost in translation. So in these cases I would think an imperfect translator would be better even though it would take some effort on the part of the recipient. The hard thing would be to find the right balance.

    Just like a calculator. It does no good if the operator does not understand the basic mathematical concept to use it correctly.

  • Ivan Malagurski says:

    I believe we should still try to learn as many languages as possibly…even if it is just to get to know the culture behind them…and also it is a great exercise for our brains :)
    Ivan Malagurski

  • Kirk McMurray says:

    I wrote a spontaneous and way-longer-than-expected comment on this article which you can read here:

    http://www.facebook.com/mcmurrak/posts/271857086257284

    • Kirk McMurray says:

      Even though I have the most transparent settings possible, FB still requires others to login in order to access the page I linked to, so I’m going to copy/past what I wrote here. That’s what I should have done in the first place, and then linked from FB to here… doh!

      >>>

      Are the days of believing that you have to (or ought to) learn a second language are over? Is it only a matter of a few years until learning a language is like learning calligraphy? Stimulating? Yes. Beautiful?. Definitely. But necessary? Not so much…

      So far the comments on the article are far from inspiring, but I’m sure a few people will chime in with convincing arguments both for and against this notion.

      My take on it. The demand for learning a second language will decrease slowly at first but in the not-too-distant future fall off a cliff. That is with the possible (or probable) exception of English. Why? Well, for two reasons:

      1. English is the most predatory language in the history of human language, so it will increasingly become THE universal language.

      2. Automated translation technologies will get better and better across multiple modalities (i.e. written, spoken) and at some point real-time translation, or automated interpretation, will become so good that you will be able to communicate very efficiently and effective across any language pair.

      While it seems logical that #2 above would make #1 moot at some point, I think the cultural, political and economic benefits of speaking English will keep the demand for learning it high for a very long time.

      Now, perhaps many of you are sitting there and thinking “this ignorant gaijin is failing to recognize that Asia is on the way up, China and Chinese will soon dominate, and the West is crumbling under massive debt!” And maybe even some of you Latin Americans think I’m a dumb gringo overlooking the rising importance of Spanish in both Americas.

      I don’t think I’m guilty of this, or at least I have a rationale as to why I think English is different. Think about it. The reward for learning Chinese or Spanish for most non-native speakers is still quite low. And while it is extremely valuable when you’re doing business in China, for instance, don’t you think there are (and will be) plenty of native Chinese speakers who speak fluent English to fulfill the vast majority of this demand?

      While I think Rosetta Stone’s revenue problems are more to do with them failing to adapt to the age of the web than with a shrinking demand for learning a foreign language, I do think there is a shrinking market for learning almost every other language than English, at least over the mid- to long-term.

      Yes, Spanish may be growing. As is Chinese, I think. We all know there are more and more non-Chinese speakers interested in learning it, or having their kids learn it. But do you really believe that demand will keep up with a) the rate at which Chinese speakers are learning English; and b) the accelerating rate at which automated translation technologies are improving?

      Now an apparent paradox is that with a combination of science, technology and the revolution of the Internet in particular, learning a language is becoming cheaper both in terms of absolute cost, but also in terms of time. Adaptive learning technologies are helping us learn things better and more quickly. (Time for quick shout out to my former colleagues at Cerego and their iKnow JP & smart.fm products, which are good examples of this).

      You can create a market apparently out of nothing by tapping into a latent demand, or even creating demand by introducing a whole new type of behavior. Perhaps someone will make it so cheap and easy and fun to learn a foreign language that it becomes more about the entertainment value of learning, or like sports, than education per se. After all, not many predicted that the demand for computer and video games was so large that it would exceed the time spent watching movies and television combined?

      But if you really want to throw a wrench in both sides of the argument, let’s take a look at the possibility of a few other ways this may play out…

      What if we are able to augment our cognition in such a way that through technology we can “learn” to speak any language (even lost languages of the past) in an instant? Think Neuroprosthetics. Think Chips implanted in our brain. Think being plugged into the “matrix” 24/7 and being merged not only with the machine, but with other conscious cyborgs.

      Or what if the biotech and genetic revolutions enable us to tweak ourselves in such a way that we all can become like “rain man” or more specifically, Daniel Tammet, who has such a unique and powerful brain that he basically learned how to speak Icelandic in a week.

      Only time will tell which of the above scenarios, if any, play out. But I am confident things will change far more quickly than most expect. As I like to say about most science fiction: it usually predicts things will change more quickly than they do over the short-term (10-50 years) but tends to underestimate how radically different things may be over the long-term (50-100/+ years).

      With the apparent geometric acceleration of change driven by innovations in science and technology, I think we need to reconsider what we mean by short-term and long-term, for making predictions that are only 5-10 years out would be like predicting 100 years out back in the 20th Century.

      Maybe we should turn this debate into a long bet. I have to think about this more carefully and try to determine the best way for measuring this. Is it the $’s spent on learning a foreign language. Is it the number of people or number of hours invested in such activity? Is a turing test for (real-time) translation? Any good ideas about a way to frame this?

      Perhaps we need an X Prize for the universal translator. The goal would be to build a device and/or application such that two people are able to have a real-time conversation with 99% comprehension and at 90% the speed at which two native speakers would talk.

    • why06 says:

      Very insightful article. I have to agree. SingHub was looking more at the narrow view of things. I was as well, but long term it will be much more likely an international language will occur, which grabs words from all languages, but I also believe it will be the democratization of language translation that causes that. As people find it easier to communicate it maybe that person learns words from 4-5 languages, and as they become more intimate with friends across the world start to learn languages.

      The universal language I think will primarily be some perversion of English, because it is the greatest linking language.

      And even beyond that its possible technology could help people learn as fast as when they were babies. (Babies learn entire languages extraordinarily fast). What I DONT think is people will just stop speaking a language. Rather I think there will be multiple languages and people will become adept at many. Each language has its own quirks and while thought sharing will break down all language barriers, I believe with the ability to learn vastly more many human beings will devour learning about human culture and communicating with people in places they never had before. Entire businesses could be built around this, with a small business hiring people in China and US all working together smoothly, what today would be a logistical nightmare for a small business just starting out will be easy then.

  • eimi says:

    Very US-centric point of view. Actually English as the world language will never happen. Is too limited and to inconvenient in its spoken form – for most people English pronounciation is a unspeakable horror.

    Also the bi- and trilingual minds are way better than just english-focused minds of Americans. You lose a culturally a lot with your lack of polyglotism, while the european nations, using often 3 and more languages in daily life, give chances to more interesting way of thinking.

  • Jared says:

    It’s good to see that this article opens the conversation on the future of language learning. Most people have not even thought about what will occur in the next few years.

    People who focus on which language will become dominate are actually completely misunderstanding the future. No language will become completely dominate. Instead instantaneous translation will allow people to learn only one language and yet communicate in tens if not hundreds of languages. Because of this the need to learn languages will almost disappear.

    A universal translator is a long way off. To translate from any one of thousands of languages to any other of thousands of languages will require time to collect enough data for this to occur. However, this is not relevant in the next few years. Only a small number of languages (less than 50) need be translated/interpreted automatically to allow the majority of the world’s population to communicate.

    Translation technology is advancing at an extremely fast pace…at an accelerating pace in fact (think of Ray Kurzweil’s accelerating returns, or Moore’s law). Iphone apps, Google Translate and machine translation programs can already instantaneously translate from one of numerous languages to another. Errors exist but we have seen over the last 50 years (longer even) that new technologies start with simple improvements and continue to evolve. In other words, the technology available today is much less capable than the technology we will see in machine translation in 5 years.

    The most common argument against this occurring is that a machine can never capture all the subtleties of a language. Perhaps not, but with billions of data points available and the initial help of human translators, machines will learn enough to allow for communication. 100% perfection is not required. 90-95% accuracy is sufficient to communicate. Just as 100% fluency is not required for people to speak a language and communicate.

    With this, a person speaking only one language will have no need to learn another language. The machine (computer, phone, earbud, etc.) will be able to manage this.

    And with that, most people will have no interest in learning a language. Our esteemed commenter Eimi, probably without realizing it, hit on an important point. Although taking a Euro-centric view and completely ignoring Africa and Asia where his comment also applies, Eimi correctly points out that many people often use 3 or more languages a day. This comes from necessity. In the United States that necessity does not exist for most people. Once the necessity for people to speak several languages is removed because of machine translation, most multilingual speakers will also stop speaking more than one language INCLUDING people in Europe, Africa and Asia.

    What will occur is a gravitation to a limited number of the most important languages worldwide, perhaps somewhere less than 40 languages. This is much more feasible in the near-term for machine translation to achieve with accuracy, than a universal translator.

    Referring back to the article, the assumption that Rosetta Stone’s drop in sales comes from less people learning language, leaves out a major factor, as Kirk correctly mentions. Namely, it ignores all of Rosetta Stone’s competitors, both paid and free. It is now possible to learn a language to fluency with free online resources and access to native speakers through numerous “pen pal” websites. Rosetta Stone definitely has not managed the transition to online learning well. THIS is the cause of their plummet more than less language learners.

  • zhit says:

    This is completly false!! Language learning is absolutly nessesary to form true bonds with people. Just imagine your in japan and u meet a hot japanese girl. Do you think you will form a relationship with this girl talking through a googl translate app? Of course not. Ive studied russian, and nothing compares to the gratification of being able to talk to people in their own language. Not only that, but it changes your view on the world and your life its self. It gives you access to a fasinating culture and people, the ability to experience the world to its fullest. So i urge everyone, instead of wasting time watching tv or playing video games, take a couple of minutes out of your day to learn a language, any language. If you need encouragement, imagine this. Remember losing your virginity? Imagine that times 10. Thats what its like to learn a language.

Singularity Hub Newsletter

Close