Who’s afraid of the Universal Translator?

Picture this: a Chinese businessman walks into a meeting room where two clients – one German and one Italian – are waiting to greet him. They are all wearing small, high-tech ear pieces which contain both a miniature speaker and a microphone. These devices are connected by Bluetooth to the businessmen’s latest generation smart phones, which nestle comfortably in their jacket pockets. The Chinese businessman greets his clients in Chinese. His smart phone picks up the sound of the greeting, converts the voice into text, and then sends the text message to his clients’ phones. Their phones receive the text greeting, translate it into German or Italian, convert the text message into voice and transmit the now spoken message into their ear pieces. It all happens instantly. The German replies in German and his response is again picked up and translated by the others’ smart phones. Each person at the meeting continues to express himself in his own language and each participant hears every utterance made by the others in his own mother tongue.

This trilingual translated conversation probably wouldn’t sound entirely natural. Only one person would be able to speak at the same time (so the translator wouldn’t be much use at raucous dinner parties) and each participant would probably need to speak rather more slowly and clearly than they would do if they were speaking to someone who shared their mother tongue. Nevertheless the speed of interaction would certainly be faster than it would have been if they had been obliged to use an interpreter; and the intelligibility of the conversation would almost certainly be better than it would have been if they had all tried to communicate in a second (or third or fourth) language.

This combination of voice recognition and instant translation is the sort of thing we are used to seeing in futuristic TV programs such as Star Trek. Douglas Adams, author of A Hitchhiker’s Guide to the Galaxy, also famously invented a creature called the Babel Fish that lives in the owner’s ear and is capable of translating any language in the universe instantly and faultlessly. While we´re still some way away from being able to travel across the universe at warp speed, (or discovering if the number 42 is indeed the meaning of life, the universe and everything) the prospect of everyone having access to a Universal Translator is distinctly less remote. Both Google and Microsoft are busy developing this technology and it seems only a matter of time before their programs become proficient. In 2014 Google announced that their instant translator was working with around 70% accuracy between English and Portuguese. Microsoft has also demonstrated a program capable of real-time spoken translation which it plans to integrate into Skype.

The key questions for anyone involved in the language teaching business are;

  1. How long will it be before such programs are both accurate and commonplace?
  2. Once such tools are widely available, will anyone still want to learn a second or third language?

The first question is difficult to answer with any degree of certainty. We all know that both voice recognition software and machine translation have a fair way to go before they can be relied upon. But they will get better, especially if giant corporations such as Google and Microsoft throw their almost limitless resources at the challenge. My best guess is that it will take rather less than 10 years …

As for the second question, we should probably separate those people who study a second language because they need to (i.e. for school, for work or to enhance their career prospects) from those people who study another language simply because they want to.

Needless to say there will always be people who want to be able to express themselves in a second or third language. There are also arguments for suggesting that fluency in another language increases brain power and might help stave off the onset of mental diseases such as dementia.

But it seems self-evident that a large proportion of the people who are currently studying a second or third language in our schools are motivated to do so largely by the extrinsic benefits it can bring, rather than by any intrinsic joy or value. A safe assumption, then, is that when a program exists that does away with the need to study another language, given the choice, a large proportion of our clients may well find some other way to spend their time and money.

This would not be good news for the language teaching business. It would be equally damaging for interpreters, publishers, exam boards, and all their myriad suppliers.

A lifeline may be provided by the world’s education ministries, which might decide to leave second or third language studies on the curriculum, irrespective of the rise of universal translators. If they continue to struggle to teach languages effectively (as they currently do in large areas of the world), private language schools should still be able to make a living by picking up the pieces. But what if future curriculum planners decide to abandon second or third languages in favour of more maths, or something relatively new like computer programming?

Before we collectively fall into despair, let’s just remember that many people in our industry (myself included) predicted that paper-based text books would have vanished by now; yet they still sell in their tens of thousands, even in fully-wired, developed countries. Could fear of the effects of the Universal Translator be equally misplaced?

The good news is that, no matter what the mid to long-term future brings, we probably still have a viable business for several years to come. At the very least, that should give us enough time to develop an exit strategy.


  1. In my personal opinion, machine translation is still a long, long way away from being reliable for a context such as the one you have described (3 businessmen trying to communicate in such a way that no misunderstandings take place). All you need to do to see its shortcomings is to copy a text of around 100 words into Google Translate and read the resulting translation. I’m sure that eventually this problem will be solved, but I don’t see that happening anywhere near the timeframe you propose. Furthermore, in order for a universal translator to evolve to the point you describe, voice recognition also needs to evolve and improve, not just to understand what is being said literally, but also to understand what is being said figuratively, what is not being said (but conveys meaning), intonation, double entendres, humour, cultural references, etc.
    I think that the only technology capable of achieving such a feat is artificial intelligence. And even though (scarily enough) several companies (Microsoft amongst them) seem to be making great advances in the field of AI, there will still be numerous legal and ethical hurdles to bypass before an AI is allowed to be “plugged in” or extensively used by humans (or at least I sure hope so myself). Moreover, such a technology is sure to be very expensive for a really long time, so I don’t really see it making a dent in our industry any time soon.

    So, I think that the language services industry will still be good for a long time to come. More so even if we take into consideration the benefits that speaking more than one language has for the health of our brains, something that I hope governments will take note of and implement into their education systems.
    I did my Master’s degree in translation in 2002 and they were already talking about the possible end of the translation industry due to advancements in software development, and yet in the year 2015, although Google Translate has done some damage due to the ignorance of many business owners on the importance of having a qualitative translation, the industry is still up and running.


  2. Hi Marco, thanks for the comment. Yes, it’s the combination of voice recognition and machine translation that we’re talking about. As I say, both Microsoft and Google claim to have made real progress in both fields – and in the combination of the two. So it’s not science fiction, or even very sophisticated AI; rather it’s a question of how long it takes to refine a system which is already operational. And the impact that development could have on our industry.


  3. Lets hope it is more complicated than it seems and it takes them longer, or we’ll all be out of a job 🙂


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.