- Posted by Stephen Whiteley
- On 31/05/2019
We probably don´t need to tell you how important language is. In a world with more than 7 billion people, language is one of the most important parts of our identities as a people. Beyond that, though, it is the most functional tool for communication.
These are interesting times, and technology has made it seem like no one has to learn new languages anymore. After all, there are all sorts of translation software packages out there to get the job done, right? Wrong!
You don’t have to take our word for it. Here are some reasons why you will be much better off with a natural translator or a translation agency other than a computerized one.
1 Non-exclusion in translation
One of the biggest problems with machine translation is that it doesn’t know what it should not translate. Some will argue that all parts of the sentence should be translated, but languages don’t work this way.
For example, the Japanese language is not so big on personal pronouns. When machine translation is trying to translate a segment into Japanese, one of the mistakes it makes is to overstate all personal pronouns, even though they don’t make any sense in the target language.
2 Translating every word
When communicating in different languages, one obvious thing is that some thoughts are implied rather than said. Thus, a single word might imply just one thing in one language, but imply lots of things in the other language.
When this happens, this sort of translation can leave the reader somewhat confused as it conveys too much, or simply something different from what was meant in the original.
3 Translating literally
When going from one language to another, the richness of the original language and culture should be preserved. That is why idioms and sayings are best translated into a similar idiom or saying in the target language.
Machine translation doesn’t normally know this and treats each word independently, translating them in such a way that they lose all cultural significance or deep meaning they might have had.
Once this meaning is lost, the context of the message is watered down and might not make as much sense as it is supposed to anymore.
4 Lack of sentence structure
Using the Japanese language as an example again, it would be difficult for machine translation to settle on the right sentence structure if it was, say, translating to English. The rules and conventions that guide where a verb, subject and the predicate can appear in English varies wildly from what is obtainable in Japanese.
Following the problems stated in #3 above, machine translation takes the easy way out and produces literal translations. This throws all the sentence elements around in such ways that you don’t even know where to start making sense of the statement from.