If computer-based technology enables one thing above all else, it’s ease of communication.
People can be reached in an instant and can use technology to bridge very real gaps in mere moments.
To say that we rely on technology isn’t an understatement; it dominates the world of today, including in business, socialising and entertainment.
Like so many other companies, DA Languages Ltd. makes use of technology every day; we send emails, make telephone calls, use the internet, and more. Our DA Link Portal and Training Centre are based online too and allow us to connect to clients and linguists from all around the UK!
Technology is undoubtedly an essential tool in our trade. However, it may surprise you to know just how big a role it plays in our industry as a whole.
What is Machine Translation?
While you’re likely familiar with the concept of translation – the process of converting text from one language into another – you may be unfamiliar with Machine Translation (MT).
Basically, Machine Translation is the act of translation, but carried out by a computer rather than a human.
You’re likely to have encountered MT in a variety of different forms. Have you ever asked Google to translate something for you? That’s an instance of machine translation. Phones and personal computers can do it too, including assistants like Apple’s Siri, Microsoft’s Cortana, or Alexa from Amazon.
As with most computer-based technology, MT is faster than the human alternative and integrating MT as a tool to be used in the workplace is already common practice.
Speed and efficiency are the main benefits of MT.
Here’s a quick look at how it works:
MT software collects data from each translation assignment fed into it. From these assignments, it builds-up a database of words and phrases that can be easily integrated into a text. The more often a system is used, the better it gets. In essence an MT system “learns” from each completed job, which then makes it easier to finish the next project. This ultimately has the benefit of saving both time and money.
That being said, there’s still plenty of debate surrounding its use in business. First and foremost is the question of reliability as no MT Software is 100% accurate.
While MT saves a lot of time by translating basic words and phrases accurately, it s often very literal; this means that nuance and subtleties of language can become confused or even get lost in translation. For example, compared to a human translator, MT can’t yet always pick up on context within a sentence. In languages like English, where homographs (words that are spelt the same but mean different things) appear often and can pose a problem. Bat is one such homograph. It can be either a winged mammal or a piece of sporting equipment. This is why human translators may often work alongside MT and in the proofreading/editing stage, can spot these errors (if they occur) and correct them.
While it can save time to use MT – especially for large documents – the human element is still essential.
Progress in Machine Translation
There are differences in the MT software that is currently available too. Basic systems (like the ones featured in search engines) won’t necessarily rival the software used by businesses, but advances in technology means that MT is always improving. Systems are designed to pick up on sentence structures and word association, rather than literal word-for-word translations. Furthermore, Artificial intelligence is playing a huge role in its development as well; if you can teach a computer to learn and make connections for itself, then translations become a lot easier and a lot faster. This means they are ‘learning’ to overcome the issues like those described above.
One example of this came as recently as November 2016. At this time, Google unveiled their Neural Machine Translation System (essentially a big improvement for Google Translate). In their tests, translations went as follows: “Japanese ⇄ English and Korean ⇄ English”. By sharing the information of these back-and-forth translations, their system made “Korean ⇄ Japanese” connections on its own, showing a logical step, but without being guided to it. Their blog post explains this with more technical detail, but the impact of this development marks significant progress for the technology.
It’s evident that over the last 10 years, MT has improved significantly. This is particularly true of languages that share common grammatical structures and patterns of formality i.e. Spanish and Portuguese. However, languages like Chinese (including Cantonese and Mandarin) still have a lot of problems due to their complex nature; they simply don’t work in the same way as languages with common similarities.
The future of translation
As a progressive, innovative company, we’ve embraced the use of MT but we are also very much on the side of the human translator. We work with a whole host of translators and promise you that this is not about to change.
At DA Languages, we firmly believe in the power of people.
We wholly trust the skills and qualities of our linguists. They’re at the heart of what we do and we simply couldn’t achieve what we do without them. You’ll see as much from a recent interview, where our translator made the following point:
“Translation will always require the human touch.”
We couldn’t agree more. To abandon such a close relationship with language to technology alone would be a shame. Technology surely has its merits, but human influence adds a unique flavour to the mix. It is, after all, how we connect. While we don’t see human translation taking a back-seat anytime soon, it’s certainly useful to be aware of the progress and influence that technology has on our industry. As more develops, we’ll keep you up to date!
Written by Rhys Pattimore
Title image produced with Snappa.com.
This is an update to an article published January 2017.