This Is Why Robot Translators Will Never Replace Humans
If you spend any time around the Web, or smartphones, or any kind of technology, you’ll know that everything is moving at a rapid rate.
Just 20 years ag
o even the idea of text-based news over the Internet seemed faintly ridiculous, let alone music, movies, television shows and video streams from relatives on the other side of the world. Meanwhile mobile phones have gone from clunky bricks to sleek, slim miniature computers, and there’s always plenty more to come.
o even the idea of text-based news over the Internet seemed faintly ridiculous, let alone music, movies, television shows and video streams from relatives on the other side of the world. Meanwhile mobile phones have gone from clunky bricks to sleek, slim miniature computers, and there’s always plenty more to come.The growth of the Web and the increasing power of the devices we carry around with us have had a significant impact on the way that we learn languages too. You can now have any text translated with the click of a button on the Web — your browser might even do the job automatically for you — while smartphone apps such as Word Lens Translator (recently acquired by Google) can even translate the text from signs in front of you in real-time.
This is just the beginning, as well: software will get smarter, networks will become more robust, hardware will get smaller and more invisible, and suddenly Douglas Adams’ idea of a Babel fish -a small creature that instantly translates speech for you- doesn’t seem so far-fetched.
In our world, though, the Babel fish won’t be an actual fish but instead a wearable gadget that uses an ever-expanding database of phrases and words for reference.
In our world, though, the Babel fish won’t be an actual fish but instead a wearable gadget that uses an ever-expanding database of phrases and words for reference.
Does this mean that human translators — and indeed teachers — will suddenly become redundant? After 
all, gadgets are always on call: they never get bored, they never make mistakes (assuming they have the right input to begin with) and they don’t expect any pay either. Are our children destined to have robot translators running on software from Google, Apple and Microsoft?
all, gadgets are always on call: they never get bored, they never make mistakes (assuming they have the right input to begin with) and they don’t expect any pay either. Are our children destined to have robot translators running on software from Google, Apple and Microsoft?In fact there are a few reasons why robot translators aren’t quite as good at their job as the human equivalents. First and foremost, so much of language and interpretation is about context, whether that’s the sentences on either side or the posture of the person doing the talking. That’s why humour sites such as Engrish.com have a never-ending stream of material to work with: a literal word-for-word translation is often the wrong one.
If you’re not convinced, try converting a few complicated sentences in Google Translate, then take the resulting text and push it back through the app in the other direction. After you’ve repeated the process a few times you could be left with something quite different from your initial paragraph (like a digital version of the old Chinese whispers game).
Translating between languages takes a great deal of skill, discernment and judgement — that’s why you can’t do it with just a foreign language dictionary.
Grammar conventions, underlying meanings, local customs, inherited understanding and other factors must all be taken into account when transferring text from one language to another. The many different versions of the Bible that are available is a useful example of this: you might be surprised at how many ways you can interpret the original Hebrew and Greek writings, and even scholars who spend their whole lives immersed in the translation of this holy text can disagree over meanings.
Grammar conventions, underlying meanings, local customs, inherited understanding and other factors must all be taken into account when transferring text from one language to another. The many different versions of the Bible that are available is a useful example of this: you might be surprised at how many ways you can interpret the original Hebrew and Greek writings, and even scholars who spend their whole lives immersed in the translation of this holy text can disagree over meanings.
Gadgets may not get bored like humans do, but they can certainly run out of battery or lose a Wi-Fi connection; they can suffer from bugs and glitches, and they’re only as good as the software programming that’s behind them — which is of course done by human hands. In the middle of an important international conference on climate change, would you want technical terms being run through a robot or translated by a professional interpreter with many years of experience and expertise?
The preferable choice is obvious.
The preferable choice is obvious.
That’s not to say software-driven translation is in any way bad — it can be a huge help in learning languages and promoting understanding across nations and cultures. But the flesh and blood translators that we rely on now are still going to be hugely important in the future, and no amount of hardware upgrades or algorithm tweaks is going to change that.
This isn’t a chess match, where the right moves can be calculated and plotted precisely by a computer — language is far more nuanced and fluid than that.
Cultural context, tone and emotion are all crucial ingredients when it comes to translating and interpreting, and humans remain superior to robots in understanding all of these areas.
Maybe one day artificial intelligence will become so advanced that it is able to translate more accurately than a human — including recognizing subtext or spotting how meanings are stringed together — but to lose the human aspect of translation would be to lose the human aspect of language, the way that it has evolved and developed over thousands of years. Language learning and translation is more than a maths puzzle, and we shouldn’t forget that.
Maybe one day artificial intelligence will become so advanced that it is able to translate more accurately than a human — including recognizing subtext or spotting how meanings are stringed together — but to lose the human aspect of translation would be to lose the human aspect of language, the way that it has evolved and developed over thousands of years. Language learning and translation is more than a maths puzzle, and we shouldn’t forget that.
About the Author
David Nield is the in-house blogger at FlashSticks. He’s also written words for some of the world’s biggest tech sites, like Gizmodo, CNET and TechRadar, as well as The Telegraph and FHM. When he’s not typing away, Dave will most likely be found catching up on an obscure foreign movie or wandering around the countryside of Northwest England.
Languages Around the Globe will always be free. However there are expenses with keeping a website up and running and devoting time and energy to provide you with more, high quality content. LATG is supported by Patreon. Click below to become a patron and earn some cool stuff for your generosity. We’re currently working to make the website advertisement free for your convenience!
Apex-editor of Languages Around the Globe, collector of linguists, regaler of history, accidental emmigrant, serial dork and English language mercenary and solutions fabricator. Potentially a necromancer. All typos are my own.
