Rage Against the Machine [Translation]
icons-action-calendar28 Apr 2016
6 minute read
While the average consumer only giggles when coming across weirdly translated phrases in Google Translate, some translators seem to be raging in anger.

Skype conversations can be translated in real time. Booking site reviews can be automatically translated depending on your language preference. Business emails written in Japanese can be rendered in French in just a couple of seconds. If you cannot read street signs in Tbilisi, Georgia, just take out your smartphone and it will guide you to your destination in whichever language you choose. Language barriers should not stop you anymore, modern technology is here to help!

Machine translation is slowly but surely establishing its presence in our everyday life. While the average consumer only giggles when coming across weirdly translated phrases in Google Translate, and otherwise might not even notice that some online content might have been translated by an MT engine, some translators seem to be raging in anger. Machines will never be able to replace the human mind, we hear translators say. MT is only used to produce cheap, poorly written texts that have nothing to do with the real meaning of translation, say others.

I looked into online translation forums to find out what professional translators have to say about the use of MT. For the purpose of this article I chose to focus my analysis on the Facebook group The League of Extraordinary Translators, a public discussion group for professional translators with more than 5000 members.

Man versus Machine

Rage against the Machine TranslationWhen browsing translation blogs, the views on MT seem to be balanced. While many authors mention that MT cannot compete with the human mind when it comes to understanding the meaning of metaphors or homonyms, for example, most of them admit that MT can offer great speed. It also has a practically unlimited memory for terminology storage, which allows translating technical, highly repetitive texts more efficiently. Steve Vlasta Vitek suggests that “the only kind of translation that may be left for me... is the kind of translation that I like to do! The kind of translation that requires an intimate knowledge of languages combined with analytical thinking and understanding of an infinite number of situations that can only be achieved by the human brain, with its billions of connections supplying the sparks needed in a mysterious thinking process.”

On the Facebook group, however, the impressions are rather gloomy. When discussing the prospects of training an online MT engine, one of its members says, “…you are sort of a supermarket cashier showing people how to use the self-checkout machines”. Somebody else adds, “Since you are going to end up having to retranslate the junk anyway, charge your usual translation rate plus a 25% premium for having to look at the junk translation and then another 25% for having to work on their platform.” Someone else even goes as far as to declare that “Post editing of machine translation (or bad translation in general) is a horrible way to die.”

Whose translation?

What we see here is that translators seem to be afraid of losing control over the translation process and the result. It comes as no surprise, since translations have been the translator’s own creation… until now. When MT comes into play, the post-editor is supposed to work with the raw output, use it as much as possible and only edit the mistakes. Giving the translation your own style is limited, if not impossible.

Another very obvious factor feeding the negativity among the translators is that hardly anyone seems to have seen good quality MT output. While many misconceptions about MT have been shaped by Google Translate — surely the most well-known, but purely statistical and far-from-perfect MT engine — even the commercially available, more sophisticated hybrid MT engines seem to disappoint at first glance. When evaluating the MT output, translators seem to automatically compare it to the human translation and identify wrong constructions or poor choice of words. After a brief look they rush to write it off completely.

Who pays the price?

Last, but not least: money. It is widely known that MT is used to speed up the translation process and to cut down on costs. The question that translators ask is – who is paying the price? While translation buyers expect translators to offer discounts for post-editing MT output, which should supposedly take less time and effort than translating the text from scratch, the translators themselves are not convinced. They often say that reading the MT output, comparing it against the source and inserting corrections (if not rewriting the sentence completely), takes just as much time as, or even longer, than translating it from scratch.

Translators have the right to be sceptical. Just like with any other kind of technological advancement, changing how we think and go about our way of working can be long and painful. To win the MT love and hate battle, one has to answer the following questions:

  1. Who is in control of the translation?
  2. What is the quality of the translation?
  3. How is the translation price defined?

Feeling lost? Come and have a look what TAUS Academy has to offer.

1 https://www.facebook.com/groups/extraordinarytranslators/

2 Reflections of a Human Translator on Machine Translation or Will MT Become the "Deus Ex Machina" Rendering Humans Obsolete in an Age When "Deus Est Machina?" http://translationjournal.net/journal/13mt.htm


Dace is a product and operations management professional with 15+ years of experience in the localization industry. Over the past 7 years, she has taken on various roles at TAUS ranging from account management to product and operations management. Since 2020 she is a member of the Executive Team and leads the strategic planning and business operations of a team of 20+ employees. She holds a Bachelor’s degree in Translation and Interpreting and a Master’s degree in Social and Cultural Anthropology.

Related Articles
icons-action-calendar27 Nov 2023
Explore the fascinating journey of Lisa Vasileva, a Machine Learning Engineer at TAUS, as she transitions from a professional translator to the field of Natural Language Processing (NLP).
icons-action-calendar1 Oct 2021
The factors that impact the reconfiguration of the translation industry in the 2020s and emerging pricing and licensing models: The Owned, Public, Private, Hosted and Shared.
icons-action-calendar3 Aug 2021
Looking into the future of the translation industry under seven sections where automated translation is no longer just a freebie on the internet, but entering the real economy of the translation sector, and it changes everything.