Machine Translation is the complex software behind such globally recognized online tools as Google Translate – but while it may seem like a recent invention, in fact it’s nearly 60 years old!
In 1954 IBM and Georgetown University researchers started developing an automated Russian to English translator by laboriously punching messages onto cards. That machine was the basis for the automated translation systems we have today – although thankfully, machine translation is now a lot quicker, and covers a lot more languages.
While those early researchers might have hoped that one day their machines would replace the need for bilingual humans to be translators, the fact is that the technology still hasn’t reached human levels of sophistication – and likely never will. There are a few reasons why. For instance, Machine Translation programs take language entirely literally, and cannot detect context, nuances of meaning, or accommodate multiple definitions. Colloquialisms and idioms that are not literal translations (such as ‘tabling a motion’) will not translate correctly, unless the program is trained to recognize each instance.
Then there’s the fact that people create new words all the time, which quickly enter the common parlance, such as ‘crowdsourcing’, ‘clicktivism’ and ‘mumpreneur’. And there’s the complexities of grammar, tense, and personal address. Many French learners are confused by when to use ‘tu’ or ‘vous’ when speaking to another, and Japanese is even more complex – if humans struggle with these linguistic tricks, imagine how a machine would cope!
These inaccuracies mean that Machine Translation is rarely suitable for use in a business context. However, it can have its uses, for instance for the translation of large quantities of non-business critical content, such as user comments on a website or product reviews – content that adds value for foreign language readers, but isn’t crucial for making a purchase decision.
We offer Machine Translation as part of our post-edited machine translation services, and can advise on when it’s suitable and will save you time and money; but one thing remains sure, and that is that human translators won’t be phased out by machines any time soon.