A Romanian company promoting its heavy industrial machinery in the German market; a Vietnamese travel company describing the stunning beauty of Halong Bay to a Japanese audience; or a Chinese company persuading a world audience to buy its solar energy panels- in the global market translations and translators are essential.
The quality of translation depends on what you want it for. If you translate simply for information, then some of the machine-based systems such as Google Translator are probably adequate. But if your aim is to publish material in another language, machine-based translators don’t do the job.
Tests of computer translators show between 50% and 65% accuracy. In other words machine-based translators usually can give readers the gist, the main points or ideas of a text. But machines are programmed to translate words; they are not programmed to understand complex grammar, the idiomatic use of language, like ‘pick up’ or nuances of vocabulary.
English, with its enormous vocabulary, is particularly difficult to translate. Take the word ‘dog’, for example. For most non-native users of English it’s a noun meaning a four legged animal descended from wolves. But it’s not that simple. The Merriam-Webster online dictionary lists 11 possible meanings for the word ‘dog’! The computer translator probably ‘knows’ 2 meanings of the word. So how well could a machine-based translation system translate this? (from a business report about an office computer networking system).
The most serious problems were mechanical. Trials of the system were dogged by breakdowns. In the main unit, for example, the central core was fastened to the main frame buy a plastic dog. The dog constantly came loose from the central core and the system stopped operating. After doggedly testing and repairing the system for a month, mechanical engineers described it as a ‘dog’.
Of course this is a made up (another idiom!) text but it shows how
a machine-based translation system would struggle to find
suitable words for these four uses of ‘dog’ and its derivatives.
And now a real translation! In the original Chinese this web page paragraph made sense but this is what Yahoo’s online translation system produced as the English version:
The company introduces United the peaceful abundant stationery industry Limited company to establish in 1997, was a fair stationery development, the production in a body's specialized company. The company since was established, gathered one group of professionals, introduces Taiwan most advanced complete set production equipment, specialized manufacture high quality folder, material book, organ package, capital feed bag and so on several series more than 200 variety work stationery products.
Apart from all of the other problems, the computer did not recognize that the words ‘abundant peace’ is the name of the company! The other difficulties include vocabulary such as ‘capital feed bag’, ‘set production’ and ‘fair stationery’; grammar- ‘to establish’ rather than ‘was established’ and non-English sentence structures. This is not the language you want splashed all over your expensive, state-of-the-art new website or colourful, glossy brochure.
The alternative is to contract a professional warm bodied human to translate your commercial documents and website. But wait! Before you rush out and Google ‘professional human translators’, beware – contracting human translators opens up another can of worms! 'Can of worms'!? Translate that! Into Dutch, no problem, ‘