As new technology continues to improve, we are offered more digital innovations to make tasks easier and more efficient. This is especially true in the world of translation. Machine translation has found a cozy spot in the industry, enabling large-scale translation initiatives to be completed more efficiently than ever before.
There are thousands of online MT tools that allow you to either speak or type text in one language and have it translated into another-- think Google Translate or Skype Translator. Maybe you want to know the Spanish word for “backpack.” Type the word into Google Translate and the Spanish equivalent, “mochila,” will appear in less than a second. Translation services providers like United Language Group have more robust MT technology that can process thousands of documents quickly, which can be a major advantage for certain large scale translation projects.
Although this technology can help expedite the translation process and save money, it has its flaws. MT is simply not as effective as human translation when it comes to deciphering complex and culturally centralized texts. Bing recently learned this the hard way when its translation feature incorrectly translated Daesh, another word for the Islamic State, from Arabic to “Saudi Arabia” in English.
The Bing error was a particularly embarrassing one, but the translation industry cannot turn a blind eye to this new technology. With that said, MT is ideally suited as a complement to human translation to ensure accuracy.
When To Use Machine Translation
Maybe the biggest benefit to MT is its ability to translate large documents quickly. MT can be useful if you’re sifting through a large volume of information you need translated fast. It’s also a great way to get the “gist” of your text.
An article in the Wall Street Journal predicts the language barrier will be effectively non-existent within 10 years thanks to advances in MT. The article’s author, Alec Ross, foresees ear buds that have the ability to translate foreign languages in real time, allowing for the smooth dissemination of ideas in any language.
New MT tools and applications for consumers keep popping up in the marketplace, and continue to be popular devices for attempting to knock down the language barrier. Even though these MTs can be handy for general translation work, they aren’t recommended without human translation as a backup. Without a trusted human source to identify cultural references and nuance, there is too much room for error when only using MT.
According to the Guardian’s article on Bing’s mistranslation, the mistake was most likely due to crowdsourced translations. In other words, if a certain number of people apply a definition to a word, even if it’s incorrect, it will appear as the “correct” translation. An open marketplace is good; but relying on users to act as linguistic auditors has its risks. Google Crowdsource, a recently released app, is also asking its users to help with translation. Hopefully we don’t see a similar blunder there.
For the time being, MT can’t pick up on linguistic nuances that human translators can. And errors, however slight, can have huge implications for certain users. When it comes to the translation of regulated documents, like legal contracts or medical device manuals, you don’t want to rely completely on machines. A mistranslated word could result in fatal consequences depending on the context.
Getting it Right
At United Language Group, we understand the importance of accuracy and timeliness when it comes to translation. We provide both human and machine translation services to our clients, and we guide you on how to apply them properly for the best results. While we embrace translation technology, our bottom line is getting an error-free document to our clients. That’s why ULG hires industry specific linguist teams that have a proven ability to produce certified translations that meet industry and client needs.
While only time will tell how far-reaching technological advances are in MT, ULG continues to rely, first and foremost, on the expertise of our human team members.