Machine Translation Will Not Replace A Specialized Translator

Hello, it has been a few days but with the Thanksgiving holiday and other commitments I was unable to write. I see that my last post is still generating traffic so I don’t want to lose momentum. These last few days I have been reading about Facebook’s translation tool through the search engine Bing and by all accounts has fallen flat in its results. This led me to write about machine translation and if it ever will replace human translation. I know that there have been many discussions on this subject but it continues to generate a buzz. Every time we mention this topic most of us grow concerned about the future of our profession and how it has hindered our position in the market. Although there have been shifts in it, there is still in my viewpoint, ways to go before machine translation can be perfected. It is true that more and more companies, agencies and other businesses are resorting to MT to reduce costs; the results of these translations are less than desirable and still need human editing. This is exactly the point of this essay. If they are to be “edited” or most of the time redone from scratch, are we then truly replaceable?

So far that has not been the case. Although we as translators have learned to use this technology as a tool it is not a replacement for our ability. I do believe that if a translation is done right, the gist, the analysis, and thought process that goes into it cannot possibly be done by a machine. Also, translation for me is somewhat of an art (we are writers after all) how you phrase, analyze and use your words are uniquely to you. For example, you may have two translations of the same text side by side, both correct and true to the original but both with their own unique style. This is what is absent in machine translated text and although it gets edited by a translator, the gist of it gets lost in the process. It is true that a machine can spit out a translation in billionths of a second and for us this would take a bit longer to produce, yet it is the end result that counts.

Given that machine translation is here and improvements to its software get more perfected, there are still fields which could never replace a specialized translator; those are law, literature, medical and marketing. These are highly specialized fields with very precise language and in literary work and marketing there is still a creative, analytical and interpretation component to the writing that is no match for a Google Translate or Yahoo Babel Fish. Actually, the more you specialize within these fields the more in demand you become. Conversely, the more generalized your specialty the more of an “editor” you are likely to become too. That is how you position yourself to become more marketable.

Finally, I would like to add from my personal experience so far, directing my service to clients rather than agencies has been of course of greater value. I use machine translation as a tool. It just won’t replace me.

11 thoughts on “Machine Translation Will Not Replace A Specialized Translator

  1. Hi Teresa,
    Thanks for the opportunity of posting my ideas on the subject.

    I once coined the quote “A translator is an author without a subject”.
    That means we translators do author written pieces of which the source subject was created by someone else, in another language and to another audience and culture. A great example of cultural differences which affect translations can be found in ‘Shakespeare in the Bush’, by Laura Bohannan – search “Saint” Google for [ translation, Laura Bohannan, “Shakespeare in the Bush”, cultural differences ]

    My homepage: says some more about that…

    Whenever someone, as an author, tries to “communicate” something to someone else, the resulting message is a combination of the author’s ideas, intentions, opinions, prejudices and emotions, including desires, neuroses and fears, all these factors being influenced by the environment, family and culture the author was raised and lived in.

    My next quote was “An author is a translator of self”…

    A translator, as a human being him/herself, should be aware of his/her own mind’s workings when receiving the original message, in order to avoid, as much as possible, contaminating the message with his/her own truths and opinions. Receiving a message from an author is such a complicated process, given all the aspects which have influenced its original creation, that any contribution introduced by a not-so-cautious translator will tend to produce a sometimes disastrous result.

    No wonder the Italians say “traduttori, traditori”.

    “Writing should be done in the same manner as the washerwomen of Alagoas practice their craft. They start with a first wash(ing), soaking the dirty clothing by the bank of the lagoon or stream; they wring the piece of clothing, soak it again, and then wring it once more. They then add indigo, soap and wring once, and then twice. Then they rinse, and soak it again, now splashing the water onto the cloth with their hands. They beat the cloth on a slab or clean stone, they wring it again and then one more time, they wring it until no water drops from the cloth anymore. Only after they have done all this do they hang the clean piece of clothing to dry, on a string or clothes line. Whoever goes into writing should do the very same thing. The word was not meant to embellish or to spark like fake gold; the word was meant to say.”
    (Graciliano Ramos, during an interview, in 1948 – my translation…)

    Back to Teresa’s point…
    I believe machines will quite soon get to the point of translating information reasonably competently, but the amount of computing power and data needed to do what our brains do is still too far in the future, if it ever gets there at all


  2. I agree with your thoughts. To think that an MT solution will someday replace human linguists is foolish. There is a time and a place when MT makes sense, but it is not a one size fits all solution. With that said, I don’t know that I am in total agreement with your comment about “fields which could never replace a specialized translator”. You mentioned law, literature, medical and marketing. With the exception of marketing, I would say that there is indeed a place for MT in these industries. I think it is more important to focus on the intended use and target audience of the translations, rather than the industry. Let’s look at the legal industry as an example. This vertical has a need to translate many different content types for many different reasons. Say a defense attorney is looking for evidence that proves their client’s innocence. Email seems like a logical place to start looking to build a case. However, can you imagine the costs that would be associated with having a human linguist translate hundreds of thousands of emails. Not to mention that these translations may not even be used if there is no valuable information found. In a situation like this, I think MT can be very beneficial. It will allow the defense to get that content translated much quicker and at a fraction of the cost. They can use MT to get the basic idea of the emails and in the event they find something that is relevant to the case, they can pay a human translator to provide a higher quality translation. The same can be said about the medical industry. By no means am I saying that a company should use MT for an IFU, but I do think it could be used for compiling the most common adverse reactions that took place during a global clinical trial. Again, I don’t see the need for human linguists going away in my life time but to say there is no place for MT in certain industries may be a bit of a stretch. So when it comes to MT, the industry is important, but the intended use and the target audience of the content is more important.


  3. True, but you can combine the both to make things more efficient for the translator. I use this online tool: and it’s made my life a lot easier. The automatic translation function is helpful for saving time, so I am not against it as long as the translation is being reviewed.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.