“It’s easy to play any musical instrument: all you have to do is touch the right key at the right time and the instrument will play itself.” It was Johann Sebastian Bach who said these words, sometime in the first part of the eighteenth century, in German.
The tricky part is, how do you know which key is the right key and which time is the right time when you are not Johann Sebastian Bach?
An organ has so many keys arranged in several rows, plus you have to remember exactly how to use the pedals and probably other things that I am not even aware of.
There are some similarities between an organ player and a translator. It’s easy to translate anything, all you have to do is supply the right words in one language for words in another language. But how do you know which words are the right ones?
Unlike musicians ancient and current, translators have a lot of help these days when they are looking for the right word in their daily search for a perfect translation, or in other words, the impossible.
I remember that it was around 1996 when I typed into Google a name of a new drug that was transliterated into katakana, one of the two Japanese alphabets used in conjunction with characters, and to my surprise, because I was close enough to the actual spelling in English, Google displayed the correct spelling for me. Foreign words in Japanese transcribed into katakana used to be one of the most difficult parts that I had to figure out when I was translating pharmaceutical validation protocols from Japanese. Japanese people like to shorten long foreign words that are difficult to pronounce and thus create new words in Japanese. For example, “semiconductor” becomes “semicon”, which is obvious enough, but “sexual harassment” becomes “sekuhara”, which is a funny word that is not so obvious when you hear it for the first time.
With Google Translate, strange katakana words in Japanese that even many native Japanese speakers don’t really understand are no longer a problem for translators. Japanese names are also a minefield because you can’t really be sure in some cases how certain geographical or personal names are pronounced. But search engines and MT programs will nowadays supply a few plausible suggestions of the right name in the right context and a human translator can then select the most likely option. When you work on a translation, you can launch Google and Google translate any time. When you play an organ, you can’t stop and look for the right key among hundreds of keys. You have to know which one to press in real time, or you can’t play.
Since just about any text in just about any language can now be simply dumped into Google Translate or another MT program, and the resulting translations often do make sense, many translators are wondering whether the appearance of omnipresent and free MT tools has lowered the threshold for expertise that will be required from translators in the 21st century.
I believe that in a way, the threshold obviously has been lowered somewhat when anything can be found on the Internet, if you know where to look and how to look. Some of the technical terms that I use in my translations are so unusual that I use them only once every few years, and then I obviously forget what they meant if I don’t see them again for several years. Human brain is designed to make connections that are stored differently in the brain depending on how badly these connections are needed. Most people remember their telephone number and even their social security number. This is information that is easily accessible because it is often needed. But most people don’t remember the phone number they had years ago if they don’t need it anymore, although they will instantly recognize it as their old number if they come across this piece of information that is still stored in a different part of their brain. That is also how “language attrition” occurs in people who are unable to speak their native language for many years.
I believe that the threshold to becoming a technical translator, for example, has been lowered, with respect to importance of correct technical terms, because anybody can find highly technical terminology in two or more languages on the Internet, or use machine translation to translate an obscure technical term.
But Internet and machine translation will not really help that much people who want to translate – if they don’t really know the language that they are translating. Google translate can suggest a translation, and some of the time, depending on the language, the subject matter and the context, perhaps even most of the time, the suggestion will be correct or at least close enough.
But how do you know that a correct translation was suggested by machine translation unless you really understand the text in the original language, as well as the subject matter in both languages?
In spite of all the help that translators have these days as their questions can be answered on the Internet almost instantly, only a translator can make this determination.
A real translation is still and always will be the product of human thinking, rather than a process in which words translated by a software program are edited by a human operator so that the result would make more sense.
That too is possibly a translation, but on a much lower level than real human translation.
The difference between real translation and MT output edited by a human is and probably always will be the difference between music played on an organ by Johann Sebastian Bach (or a lesser player, depending on the translator), and music played by an orchestrion, which almost sounds like real music played by a whole orchestra.