Why computer generated translation is not the best?
Automated Translation
Advancements in artificial intelligence have led to a stage where we have a variety of computer software available with us for translations. These are especially useful when people have to understand the basic synopsis or meaning of big documents in another language. To a great extent, if made use properly, these software are also beneficial to translators and language experts. However, all such software vary in terms of quality of the translation, and while translating from Spanish to Italian may be quite accurate because of the similarity in both languages, it may not be the case from German to Chinese.
If one goes by understanding a language in terms of the context, or the peculiarity of it, depending in various idioms etc., then such software can be considered somewhat poorly equipped. In such cases, a good translator, looking for preciseness will not be able to benefit much. For instance, legal documents and others on science and technology have various specific terminologies, which may hold very different meanings in different languages.
Moreover, the grammar rules along with sentence structuring and construction are different in all languages. Thus basic rules of one language may not be applicable in the same manner to the other languages. Besides, due to changes in the range and variety of words in relation to the vocabulary of each language, it becomes difficult at times to find out the exact words. For “like” and “love” both, the French translation comes out to be “aimer”, and thus where a person is trying to differentiate between the two, there is a difficulty. This kind of a problem occurs with other languages too, and not just English and French.
For one sentence, if a person uses four different translation software or translation engines, it is most likely to happen, that he/she might get at least 2 different versions of the same statement, if not exactly 4. It becomes increasingly difficult to understand whether the translation is catching the intensity of the word and its meaning or not. In case it is not, then the purpose of exact and good translation gets defeated.
Also, for those who are frequent users of such translation engines online will know that most of the translations are completely nonsensical. For every language, and each word that is included into it, there is a cultural context, and the habits, customs and behavioural patterns of people attached to it. This is something that machine translations cannot grasp.
Being a translator this is my point of view, but Im sure all you readers out there surely have something different to say. Have you ever come across an instance where the online/software translation was utterly rubbish and funny? I am sure everyone must have had such experiences.
No matter how much they develop and improve computer-generated translation software, in my opinion it will never be able to replace the human translator. Sure, this software can assist in gathering an idea about what a text is about, although have you ever tried to translate a website with Google – you end up with 90% of the translation consisting of the source language.
Until they can teach computers to think, judge contexts and observe subtleties, humans will still be a very necessary part of translation.
Hi Lia…… I totally agree to what you have written.
Here is the ultimate automated translation example:
http://boingboing.net/2008/07/15/chinese-restaurant-c.html
Fresh example. I’m translating specifications for pulsers-receivers from English to Russian.
Looking for a best translation for “abrupt voltage transition” I entered it in Google. Navigating through the research results I found my phrase on the Google Translation page with the litteral translation: “abrupt voltage of the transition”.
I wonder why have not the English-speaking Google translation machine devellopers taken into consideration the elementary English phrase construction rules.
The ultimate challenge is finding the perfect combination of machine translation and human translation.
Yes Ronald, you are right on this point a machine will never replace a human translator.
On the other side by combining paper dictionaries with Babelfish + internet dictionaries one can do any language’s translation. Without any problem. You are getting this way more time to refine your translations.
The different terminologies, even in the same language can make problems: i.e. in French the judicial language used in France differs completely from the one used in Belgium/Luxemburg, Switzerland or Quebec. To tighten the nut on us poor translators the judicial language differs when speaking about general texts or immobiliar contracts.
Sometimes the court decisions one have to translate comport synonims one can hardly translate as their meaning is simply the same – I am well against this sort of “word diarrhea” the courts are accustomed to.
Last but not least: the best way to do a good translation lays not so much in the knowledge of language, but a lot more in knowing the logic of the speakers – simply said the “spirit of the spoken tongue”. Then you can translate the most extreme languages, like Persian and Estonian with only 30-60 expressions, after a prepartion of 15 to 30 minutes.
It is incredible but it works !
Advancements in artificial intelligence have led to a stage where we have a variety of computer software available with us for translations.
My attitude towards MT used to be venomous. However, I have since changed my mind to some extent. The way I see it is this:
#It’s fine for ‘gisting’ short messages, when speed is of the essence
#It has made quantum leaps in quality. One example is the SDL version which handles DE>EN very efficiently, despite the grammatical differences. What is crucial is having a dictionary that is tailored to the individual customer’s requirements.
#Either way, there were will always be work for humans in ‘post-editing’ MT text
#It might take over to a larger extent in the future, but when? Maybe when the first real Starship Enterprise is launched?
#I cannot understand how text with subtle nuances such as marketing and literary stuff could ever be handled by MT
As Attila rightly says “The spirit of the spoken togue”.
is there a good free french translation tool on the internet ?”:,
I am really thankful to this topic because it really gives useful information ;:,
Leave your response!
Follow Us
Fatal error: Call to undefined function curl_init() in C:\Inetpub\vhosts\translationartwork.com\subdomains\blog\httpdocs\wordpress\wp-content\plugins\dandyid-services\class.dandyid.php on line 68