As the business has become as international as it had never been before, there is a growing demand for translations in numerous fields. Companies, businesses and individuals in need of a translation look for a good job that is done in no time and with the best quality. In this regard, machine translation tools (MT tools) come into the game which has been more and more developed over the past few years. There are many major translation companies worldwide that are interested in expanding their output through these tools. But are these machine translation tools as reliable as praised, and could they replace the traditional translator in the future?
For translators, they are the enemy. For the translation industry, they are the solution of tomorrow. Machine translation tools or simply MT tools are an intensely discussed topic coming along with many benefits as well as drawbacks. Naturally, the biggest advantage is that MT tools translate any kind of texts within seconds. The biggest disadvantage, on the other hand, is that it can never be said for sure if the target translated is grammatically and verbally correct.
Every language has its own vocabulary, grammar and syntax which Machine Translation engineers try to integrate into the algorithm of machine translation tools as best as possible. The important task of the MT tools is then to translate a source text from one language correctly into the target text of another language. And that is where the drawbacks often show up. Either the algorithm has not been elaborated sufficiently to grasp the correct sentence structure or the tool translates a certain word with the wrong term.
Let’s take English and German as an example. English is the language in the world with the richest vocabulary with over 600,000 words according to the Oxford English Dictionary (however, there are probably only about 170,000 words in current use).
German comprises a richness of vocabulary as well but, in contrast to English, it currently only amounts to about 350,000 words. Enter any given German word into an EN>DE online dictionary and in a blink of an eye, you have vaguely 20 to 25 English terms for one and the same German word. So it goes without saying that there are no machine translation tools in the world yet that can comprehend completely the sense and spirit of a text in order to translate it perfectly into the target language.
In the field of technical translation, medical translation and legal translation, as well as other areas, mistakes can cost a translation company perhaps clients and its good reputation. It is unacceptable when there are wrong terms in a crucial text or when a native speaker finds himself in a situation where a text is hard to read because of the wrong syntax. Mistranslations could even lead to drastic problems, such as for example in the field of medical translations.
Currently, no machine-translated text could ever leave a translation company, that places value on good quality, without being checked by a translator or proofreader. The amount of work in the area of machine translation post-editing (MTPE) is increasing. Sometimes, however, post-editors face such bad machine-translated texts that the translation might have been finished faster if the translator had simply translated the text in a conventional way. That’s why it is in some cases hard to say if machine translation tools speed things up or not.
Summarising, machine translation tools have changed the modern translation industry and it is very unlikely that they will disappear from the landscape. Nevertheless, human translators cannot and should not be replaced by a machine, as a well-translated text relies on the eye, the passion and the creativity of the translator, who shows his/her love for language through his/her great work.