Reports last week claimed that 40% of jobs would be replaced by machines by 2030, and that they will be able to ‘translate and interpret text quicker than humans’.
Many companies already use machine translation to provide quick and free translations of their websites and other materials, so it is down to us as language service providers along with our team of trusty translators to explain the added value of human translation.
But where do we start explaining to a company with their eye on the bottom line why they should invest in professional translation? Here are a few of our suggestions:
Take the Spanish idiom ‘Estar más sano que una manzana’, for example. Google Translate tells us that this means ‘Being healthier than an apple’. While this is literally correct, the equivalent English idiom would be ‘to be as fit as a fiddle’.
Only a human can have the required sociocultural knowledge and research ability to localise such ‘fiddly’ phrases (excuse the pun!).
Everyone has heard that Coca-Cola story by now. In Chinese, for example, trying to create both the correct phonetic sound of a brand name and use symbols with appropriate meanings is a bit of a minefield that only human specialists can navigate.
Giving contact details to a UK based phone-line on a French website? A human will add the +44 dialling code, but Google Translate won’t.
Directing Chinese customers in mainland China to your Facebook page or YouTube video? You can’t – they’re blocked. A human would be able to warn you of this; a machine wouldn’t.
The EU has strict regulations about the labelling of food products, such as the inclusion of allergen information that could be potentially lifesaving. Do you really want to risk a lawsuit by leaving this in the hands of a machine? We know that humans make mistakes too, but here at Web-Translations we always include proofreading by a second linguist for precisely this reason.
And remember, online machine translation stores all of the text that is put into it, which can raise serious concerns around confidentiality. While you can’t ask a machine to sign an NDA, a human translator will.
As a native speaker of a language, you are an expert in understanding its quirks and peculiarities.
European French generally precedes colons, semicolons, exclamation and question marks, with a non-breaking space. A human translator wouldn’t have to think about this twice, but Google didn’t think about it at all in the example above.
Meanwhile, other Francophone nations wouldn’t all abide by the same conventions, but as Google Translate doesn’t differentiate by locale, it makes the assumption that each language is the same regardless of location. This is particularly troublesome when dealing with variants as diverse as European and Brazilian Portuguese, where a train is a ‘trem’ in Brazil but a ‘comboio’ in Portugal.