Machine translation: is human translation under threat?

In recent years, the technology responsible for machine translation has improved dramatically, mirroring other industries and raising the question of how likely machines are to replace human labour.

Machine translation can take a lot of the arduous work out of more simple or repetitive translations and hugely speed up the process. Instruction manuals are an excellent example of where these computer algorithms can be used to provide a fast, reliable, and accurate translation where a human translator working alone may not have been as efficient.

And it’s not just within the translation industry that machine translation engines and Computer-Aided Translation (CAT) tools have become ubiquitous. Tourists often rely on mobile applications or websites to provide translations of important signage they might come across on their travels, or even to make basic conversation with locals. In contexts such as these, machine translation proves useful, as it offers speed, convenience, ease-of-use, and allows the user to communicate effectively where they previously may not have been able to.

However, when it comes to more complex translation situations, where the stakes are higher, in a business, legal, medical, or political context, for example, machine translation may not be the better option. Especially where words have more than one meaning, which is often the case in legalese, using machine translation can lead to crucial errors and detrimental misunderstandings. In contexts such as these, the “hit and miss” nature of machine translation means that the human component is still needed, if not to translate the content in its entirety then at least to proof-read, edit, eliminate mistakes, and optimise the end result.

Many also claim that CAT tools actually impede creativity, and that where a text aims to create a relationship with the reader, the human component remains essential in order to make the translated text sound natural and capture the true sentiment of the original. Despite all the recent technological progress, computers still lack the ability to properly understand and express intonation, emotion, irony, sarcasm, poetic nuance, cultural references, figurative language, and other such key, yet subtle, aspects of human communication.

In brief, human translation is only under threat if machine translation stands a chance of being able to completely replace it. With the current limits on machine learning, human translation is still the higher quality option. Of course, machine translation may be favourable to some employers/industries as there will be lower costs with less human wages to pay; but where the quality and accuracy of the translation is the priority, human translation should still be the norm.

For the moment, machine translation can definitely help and be used in conjunction with the skill of a human translator, but without superseding it. Of course, machine translation will continue to develop all the time and may become a real threat in the future, but until then, it should not be considered a total replacement for its human counterpart.

Photo by Simon Abrams on Unsplash