23 October, 2017
The use of various versions of computer translation programs is becoming more and more common. The customers and providers of translation services and translation agencies are paying more and more attention to the opportunities offered by up-to-date information technology.
Nevertheless, with all the variety of proposals, the use of software for translations from and into a foreign language is still not clear. A person who is not related to translations, as a rule, does not see differences between two types of computer programs used for translations. But they differ from each other significantly, and, more specifically, each one is designed to solve certain tasks with its own limitations.
This term means that the translation of any text (at present mostly written, and in perspective - oral) is done with the help of a special computer program.
Unsophisticated user believes that in order to obtain an adequate translation using such program, it is enough to download the material in a foreign language. Many people are convinced that after minor proofreading they can have a correct translation. This point of view does not reflect the actual situation.
Computer-based translation programs deal with words, and not with meaning. In most cases, the text received as a result of the computer-based translation cannot be edited quickly and is even hardly understandable. Machine translation programs do a good job for short simple sentences, but when translating more complex constructions they can not convey meaning as a lot of words, for example, in English, have several meanings, and the computer algorithm chooses only one of them.
Anyone who used, for one reason or another, even the most up-to-date machine translation program, realized fully that the result was not a "preliminary" version of the text, but a primitive "metaphrase." The computer produces a set of linguistic structures, simply replacing the words of the source text with the corresponding words of the desired language. In this case the meaning can be distorted completely. For sure, computer-based translation programs are developing actively. Especially it concerns the programs intended for on-line translations: the developers use various algorithms of "feedback", track user corrections and improve the programs accordingly.
There are language pairs for which computer-based translation is done more or less satisfactorily. Machine translation programs have their own sphere of application. They can be used to process large arrays of texts. For example, to get a general idea of the text or search for the required extract in the text.
Another advantage of machine translation programs is that the computer can operate continuously.
These programs are called Computer-Aided Translation (CAT). It is a specialized software designed to facilitate translators’ work.
While using CAT, the translation is carried out by a person, not by a computer, and the program performs an auxiliary role. CAT programs facilitate the work of the linguist and improve the quality of the finished translation. Therefore, most translation agencies require their employees to use Computer-Aided Translation.
These software tools store information about previously done translations, make thematic glossaries, use algorithms to select more relevant linguistic structures from previous texts. But such programs are absolutely not suitable for an independent use by people who have small knowledge of foreign language.
In practice of translation agencies there are customers who try to lower the cost of translation services by providing a translator not with the original text, but with the text translated by the computer program. Such customers expect that editing of the computer translated text will be cheaper.
As experience shows this option creates a lot of work for a translator, and the result is less qualitative. The cost of work in this case, naturally, becomes higher.