The Universal Translator: Fact or Fiction?

Speech bubbles with different flags on blue background, seamless pattern

Is the prospect of being able to communicate with anyone in any part of the world using an instant universal translator something that is remotely possible or is it just science fiction? If you are a NAATI translator, you probably think that any sort of really useful machine translation is years away from happening, at least for the sorts of jobs that a professional translator has to deal with. But the potential of a universal translator to transform the level of global communication is huge, so don’t be surprised that the advent of such a machine is not as far-fetched as one might think.

For many individuals and businesses, translation must seem an unnecessarily slow and time consuming obstacle in the way of being able to get a message across borders, and not just across borders, too. Think of the modern western nation with its increasingly cosmopolitan population of migrants and tourists, let alone the plethora of global business interests clamouring for a place in the market. Wouldn’t it be so much easier if all these people that make up an individual nation could communicate with each other quickly and easily?

At present, machine translation tools like Google Translate and others like it are not really up to efficient and accurate communication across language barriers. Even so, machine translation has come a long way from its beginnings in the 50s. The early translation tools used word for word translation with a built in dictionary. The tools used today are a lot more sophisticated and getting better all the time, with context based translation and statistical tools used, but machine translation would never pass muster with the sorts of customers that use NAATI translation and its equivalent in other parts of the world other than Australia.

The sometimes faltering steps towards a universal translator are often derided by observers as they often seem to fall back on a human translator, even if it is as a final step in a machine translation process. That’s been the experience, for instance, with the integration between machine translation and crowdsourcing. Some say that a universal translator will never be able to fully understand all the nuances of an individual language, with the cultural intricacies, dialectical variations and idiomatic expressions, but with incremental steps it’s getting closer!