We recently posted an amusing article on our facebook page highlighting the pitfalls of poor translation. These mishaps were most likely the result of a quick tap into an online translation site followed by a hasty cut and paste. But as the world becomes an ever smaller place and the internet never more than a click away, the temptation to take these short cuts increases. But beware; a poor translation of a restaurant menu can be the source of a few laughs but for many businesses, it could prove not only embarassing but costly. For example, Coors, the American beer brand, spent millions of dollars rectifying their "Turn it loose' slogan for the Spanish market which they later learnt that when translated, was commonly interpreted as 'suffer from diarrhea'- a problem that a native translator would have avoided immediately no doubt.
With global markets increasingly easier to access via the internet, our need for effective communication is greater than ever as we export goods and services to all corners of the world. To aid communication, we have been inundated with various programmes and apps promising to connect us to the world in all of its 6 500 languages; the latest being a 'Universal Translator' app that claims to use speech synthesis and voice recognision to translate 'any language from any language'. So has this made language learning obsolete? Not quite yet.
Online translation tools are programmed with a language's grammar rules and vocabulary. When text is applied, the translation device applies these rules and provides a rough translation. More recently, online translation programmes have worked using a giant database of human translations to try and mesh together a more natural sounding response. This is called 'statistical machine translation' and though it is definitely an improvement on the previous method, it remains far from perfect. They lack the ability to interpret tone, register and context. They struggle when it comes to metaphores and idiomatic expression. They can't cope with human errors like spelling mistakes or understand subtle nuance. All of these factors prevent a computer performing the job of a linguist. Languages are a complex mixture of components that can't simply be plugged into an algorithm with any measure of accuracy.
When it comes to business, it is definitely cheaper in the short term to translate a website or slogan via an online translation tool but grammatical errors, poor sentence structure and sentences of complete gibberish are to be expected. Translation agencies that employ experienced, native speaker translators ensure all blushes are spared and professionalism is maintained. They are also aware of the importance of localisation. Languages are not homogenous; what may be perfectly acceptable in one region may convey a completely different message in another. One Airline company experienced this issue when promoting it's use of leather seats on-board. Their slogan, 'Fly in leather' translated well into Spanish but in Mexico, 'Veula en Cuero' was interpreted as 'Flly naked'. Not quite what the company was advertising...
So despite constant advances in the sophistication of technology, it would be safe to say that we are currently a long way off technology replacing the value of linguistis and the art of language learning.
With global markets increasingly easier to access via the internet, our need for effective communication is greater than ever as we export goods and services to all corners of the world. To aid communication, we have been inundated with various programmes and apps promising to connect us to the world in all of its 6 500 languages; the latest being a 'Universal Translator' app that claims to use speech synthesis and voice recognision to translate 'any language from any language'. So has this made language learning obsolete? Not quite yet.
Online translation tools are programmed with a language's grammar rules and vocabulary. When text is applied, the translation device applies these rules and provides a rough translation. More recently, online translation programmes have worked using a giant database of human translations to try and mesh together a more natural sounding response. This is called 'statistical machine translation' and though it is definitely an improvement on the previous method, it remains far from perfect. They lack the ability to interpret tone, register and context. They struggle when it comes to metaphores and idiomatic expression. They can't cope with human errors like spelling mistakes or understand subtle nuance. All of these factors prevent a computer performing the job of a linguist. Languages are a complex mixture of components that can't simply be plugged into an algorithm with any measure of accuracy.
When it comes to business, it is definitely cheaper in the short term to translate a website or slogan via an online translation tool but grammatical errors, poor sentence structure and sentences of complete gibberish are to be expected. Translation agencies that employ experienced, native speaker translators ensure all blushes are spared and professionalism is maintained. They are also aware of the importance of localisation. Languages are not homogenous; what may be perfectly acceptable in one region may convey a completely different message in another. One Airline company experienced this issue when promoting it's use of leather seats on-board. Their slogan, 'Fly in leather' translated well into Spanish but in Mexico, 'Veula en Cuero' was interpreted as 'Flly naked'. Not quite what the company was advertising...
So despite constant advances in the sophistication of technology, it would be safe to say that we are currently a long way off technology replacing the value of linguistis and the art of language learning.
RSS Feed
