The crew of the Starship Enterprise have the Universal Translator. The Doctor’s TARDIS uses a telepathic field to provide instantaneous cross-species translation. Douglas Adams provided a slightly grosser solution to pan-galactic communication in The Hitchiker’s Guide To The Galaxy with the ‘babel fish’ – a small leech-like fish that was inserted into the ear (and which gave its name in real life to the now defunct Yahoo! Babel Fish online translation platform).
Sci-fi writers have always envisaged a day when instant, flawless translation is not only possible but commonplace. How close are we to that goal in 2015 and what would it mean for the translation industry?
The core of most translation companies is built around human translators, but that doesn’t mean translation professionals don’t use systems and software that significantly aid the process. For large or ongoing projects, translation memories and glossaries can be invaluable. These can store, compare and re-use previously translated content greatly reducing time and expense and ensuring consistency. Technical and business terminology used within an organisation can also be stored for future use.
Organisations looking for on demand machine translation may also be able to plug into an API. This essentially provides a direct link from your content management system (CMS) to the translation provider’s, giving you access to quick and easy translations. This can be great for tasks such as keeping multilingual websites up to date, sending blog posts for immediate translation and understanding queries and emails from overseas customers. Machine translation (MT) can be combined with professional editing however to cast a human eye over the material and produce fluent, accurate results in a fraction of the usual time.
Not everyone has access to bespoke translation software of course but some rather big names remain determined to bring high-quality, convenient translation technology to the masses.
Google is leading the way, claiming that more than 500 million people use Google Translate every month, making over a billion translations every single day. But what sort of translations are they making? Most of us are familiar with the basic web interface, which allows you to enter text and translate it at the push of a button. This sort of automatic translation has certainly improved but straight dictionary translation is always prone to errors. Try pasting a chunk of text into Google Translate, translating to a random language then back to English to see what we mean.
At the start of the year Google also announced a couple of major updates for the mobile Android and iOS Translate apps. The recently acquired Word Lens technology allows for instant translation of text. Simply point your device camera at a menu, sign or other text and the translation will be overlaid in an ‘augmented reality’ format. This is also planned to be incorporated into Google Glass when the wearable tech finally launches. Spoken translation apps are not new but Google Translate has sped the process up, with automatic recognition of the languages being spoken.
Google’s big rival in all this comes in the shape of Skype – which was bought out by Microsoft in 2011 for a reported $8.5 billion. Last month it removed the sign-up process to its Skype Translator preview, effectively unveiling a ‘real-time’ translation process that can present written translations of verbal communications to each participant in a conversation. This could be incredibly useful but again there are limitations. Some previews revealed the ‘real-time’ processing could actually take around 10 seconds a time, leading to stilted conversations. It struggled with regional accents and is so far limited to just a handful of languages.
Improvements to consumer-centred services like Google Translate’s mobile app are obviously great for anyone looking to order tapas on holiday or find directions to an unfamiliar foreign destination. Skype will no doubt be targeting business with its vision of video-conferencing without linguistic barriers.
Some governmental bodies regularly use machine translation, but there are usually caveats. The US Deaprtment of Labor said last year: “It is seldom, if ever, sufficient to use machine translation without having a human who is trained in translation available to review and correct the translation to ensure that it is conveying the intended message.”
Businesses too should be wary of over-reliance on machine translation. Taco Bell embarrassed itself in Japan recently when a relaunched website promised ‘low quality chips’ and ‘Supreme Court beef’ on the menu.
Technology is improving all the time but for now it needs to work in conjunction with human translators for the best results.