With machine translation on the rise, will translators be needed in the future?
Have you used machine translation recently? The speed and quality of today’s neural translation software is certainly impressive, but when words matter, the result will often sound a bit off. So what are the roles of humans and machines in the future of the translation industry?
Photo credit: Lyman Gerona
The need for translation services has grown proportionally to the expansion of commercial networks and activities. The need to clearly communicate with teams and clients across national and linguistic boundaries has made translation services indispensable for international companies. In parallel, the rapid development of machine translation technology offers companies a quick and cheap alternative to traditional translation services. But how does machine translation hold up against human translators?
How does machine translation work?
Even though machine translation has been around since the 1950s, this technology has developed at breakneck speed in recent years. There are different machine translation systems, though the most advanced and the one that is producing the best results is neural machine translation. Neural machine translation teaches software to translate using machine learning technology based on artificial networks inspired by central nervous systems formed by countless, interconnected small units. The machine then processes and analyses massive volumes of existing data to determine not only translations, but also patterns and structures.
There are numerous providers of machine translation, and many of them are free, such as Google Translate or DeepL. Tech giants Google and Microsoft have embraced and introduced this technology. Amazon has its own neural machine translation service and Apple announced that their new operating system, iOS 14, will be equipped with a translation app.
Keep in mind that sensitive information and personal data should not be entered into translation machines, especially if the service is free of charge.
Considering the short amount of time that neural machine translation has been around, its development really is staggering. Machine translation technology is advancing at a very rapid pace despite still being in its infancy. Furthermore, it is already changing the translation industry as we know it, as many are opting to use machine translation and then post-edit the texts it produces. This makes many human translators fear for their jobs, but should it?
When should you use translation software?
There are advantages to machine translation in terms of speed and cost – it’s instant and usually free. And even though machine translation is revolutionising the translation industry, it will by no means make human translators obsolete any time soon. Machine translation is not adequate for all types of texts – especially those involving a more elaborate, literary or idiomatic use of language. The results produced by machine translation are often not ready for publishing and still have to be checked by a human editor. Although the technology is still being developed – and at an extraordinary pace – machine translation is missing something that only human translators can offer: Sprachgefühl.
Why machines cannot replace human translators and editors
Computers were designed to solve mathematical problems, but producing an accurate translation is a different ball game. Translations cannot be approached with the same mathematical principle, because simply translating words directly does not guarantee accuracy. In fact, the inability of engines such as Google Translate to accurately translate content often ends up as the butt of jokes on Twitter and Instagram.
Translating is more like a kind of writing, and writing is an art form. That is the difference between professional human translators and machines. Computers do not understand context or culture and are unable to play with words and syntax. Making sense of a series of words arranged into sentences building paragraphs requires not only knowledge of a language but cultural sensitivity and familiarity with particular contexts, idiomatic expressions, tone and style and, in the case of companies, corporate language and identity.
The future of translation
Using machine translation might be useful for quick, non-vital, low-priority content, but corporate communication that adheres to the particular identity of a company should be translated by human translators. As far as post-editing goes, we would contend that the advantages are not really substantial unless you are prioritising speed over linguistic quality. Our editors can translate a source text faster and to a higher standard compared to post-editing machine-translated content.
Make no mistake: machine translation is not going anywhere, but neither are human translators. What a time to be alive!