Machine translations from Facebook to Bing are improving all the time, but the shortcomings of the technology “are inevitable”. Indeed, anyone thinking they can run text through Google Translate and pass it off as professional is very much mistaken.

Such is the opinion of English as a foreign language lecturer Anne Merritt, who writes in the Telegraph that she – like all language teachers – is always able to tell when her students have put their essays through an automatic translator rather than do the work themselves. In addition to simply containing mistakes, the finished article is often “incomprehensible”, she says.

Not only can machine translations be confusing, but they can also be offensive, as numerous advertising disasters have proven. Ms Merritt points to Pepsi’s ‘Come Alive’ slogan being translated for its Chinese market to “we bring your ancestors back from the grave”, while other examples include Nancy’s Petite Quiche advert, which included the slogan: “Petite bites. Big compliments.” Unfortunately, in France “petite bites” means something quite different to what was intended. Certainly, these are examples that prove no marketing department should ever rely on a machine translation – or even one done by a human who is not a professional translator – if they want to make the right impression.

The main reason machine translations are not yet in a place where they can compete with a human is that syntax proves a stumbling block. Syntax is what makes language human – while apes can be taught to ask basic questions or make demands, it is syntax that makes these words a sentence. Therefore, without syntax, it is impossible to create a well-written and comprehensible text. However, syntax remains a real challenge for machine translators, with Ms Merritt noting the technology is not “sophisticated” enough to deal with it correctly.

Yet there is a place for machine translation, she claims. When someone needs a quick translation of a recipe, route directions or social media updates, they are more reliable because the length of the text is generally shorter. Thanks to social media networks like Twitter and Facebook providing their own machine translators, people can communicate with other users all over the world and speaking a different language no longer has to be a barrier. This looks to be a trend that will only continue.

It seems that, just like the search engines, social media platforms are investing in translation technology. Earlier in August, Facebook announced it had acquired Mobile Technologies; the start up behind online voice translator app Jibbigo. Users of the app can record their voice or write text and then translate it into their choice of 25 languages. Tom Stocky, director of product management at the site, said: “Although more than a billion people around the world already use Facebook every month, we are always looking for ways to help connect the rest of the world as well.” He added that he viewed voice recognition software as an important part of this ambition and concluded that Facebook regarded the acquisition as part of its “long-term product roadmap”.

With developments like this coming thick and fast, the universal translator may soon be within grasp. However, Merritt warns: “When working in foreign language, it is not simply translation that brings meaning to a text, but also cultural subtleties, double meanings, and local slang. Until a [reliable machine translator] hits the market, translation software should be used with caution.”