In the past, we’ve spoken at length about machine translation. The technology has come along in leaps and bounds over the last few years, and while it’s unlikely ever to replace human translators outright, many now reckon it has a role to play in the language professional’s workflow.
But what about machine interpreting? The notion of two parties being able to communicate in separate languages without a human intermediary has long been the stuff of science fiction – you might recall Star Trek’s Universal Translator, which ostensibly scanned brain-wave frequencies, or Douglas Adams’ miraculous Babel Fish from the Hitchhiker’s Guide to the Galaxy. But nowadays, surely both speech recognition and machine translation software are fast and powerful enough to provide a real-world equivalent?
The state of machine interpreting today
There have, in fact, been a few developments in this space recently. There’s at least one machine interpreting solution on the market already, with Israeli startup Lexifone having launched a telephone-based service in 2013. What’s more, researchers from the Nara Institute of Science and Technology are understood to be working on a lag-free interpreting system for the 2020 Tokyo Olympics, which will reportedly transpose the games’ Japanese commentary into English in real-time.
Even technology titan Microsoft is getting involved. Last month, the company announced a new feature for the ubiquitous VoIP and video-conferencing program Skype, which it acquired in 2011. Called Skype Translator, the new tool – expected to enter public beta-testing this year – combines speech recognition, automatic translation and artificial intelligence to provide a supposedly Star Trek-style machine interpreting experience.
All in all, there’s little doubt that speech recognition is getting more advanced, or that computers are growing fast enough to produce sophisticated machine translations in split seconds. But will these innovations find mass acceptance in the business world, or will they come up against obstacles to accuracy that render them unsuitable for anything but a handful of niche tasks?
The problems with machine interpreting
Unfortunately, the latter will probably turn out to be the case. The most basic reason for this is that it relies upon machine translation, which has come to be regarded as a useful tool in the language professional’s arsenal, but by no means a replacement to a real person’s attention to detail and cultural insight. In machine interpreting, this is only one link in the chain – it also requires speech recognition, which as anyone who’s ever used a modern smartphone’s voice commands will tell you, has a long way yet to go in terms of accuracy.
Another thing to remember is that an interpreter is far more than just a person who turns up for a job, listens to one individual speak, then renders that sentence into another language. A good interpreter has to have an in-depth, up-to-date understanding of a language’s quirks, nuances and colloquialisms, as well as the way its speakers prefer to conduct business. He or she also needs excellent interpersonal skills, in order to help banish any cultural barriers that might impede the conversation.
This is to say nothing of body language. Interpreters are regularly expected to extend their services to face-to-face meetings, during which not all communication will be verbal. By picking up on conscious and subconscious gestures, expressions and tics, a talented interpreting professional can bring a whole new level of mutual understanding to a multilingual dialogue.
Finally, consider the current state of computerised speech. Real-time machine interpreting solutions are required not only to transpose speech from one language to another, but also to provide a verbal output. Granted, speech synthesis is far better now than it was a few years ago, but still often falls short of acceptable standards in pronunciation, emphasis and tone of voice. Listening to computerised speech for a prolonged period of time can be exhausting – imagine having that experience exacerbated by an inaccurate translation!
Of course, the machine interpreting market is a fledgling one and there’s no saying what the next wave of technological innovation will bring. For now, though, it’s probably safe to say interpreters will remain in demand for the foreseeable future.