Syntactic Challenges of Using a Translation Memory Database

What happens when a 128-page technical text needs to be translated in four days? In such situations, no translator would even consider working without the aid of translation software.

Advantages of Translation Software

If you don’t use any type of translation software, you’d have to translate the entire 128 pages mentioned. But if you did, you would quickly realize that the text contains repetitions, which only need to be translated once, and you’d end up translating about 52 pages instead.

Today’s translators and translation offices in particular, definitely like using translation software. More importantly, clients love them. Translation software allows human translators to save time, save energy, maintain quality throughout a file and, in the end, saves money for the client, too. It would seem that everybody wins!

Challenges with Translation Software

First, it takes time to learn to use the tool. The software is generally chosen by the client, which means it is in your own interest to become familiar with more than one type of translation software. Be open to learning about new software, and make the effort to remain up-to-date on translation tools and their benefits.

Secondly, you will likely encounter some sort of technical problem while using translation software, since technology is never perfect. Be prepared mentally.

Syntactic Challenges of Translation Memory Database

When it comes to syntax, there will be challenges too. For example, Turkish syntax is very different from English syntax. Long sentences divided into three different segments are problematic for Turkish translators. As a native speaker, let me illustrate why:

Say I’m using a translation tool for this 128-page file. I’ve come across a sentence divided into three different segments while translating the fourth page:


Here’s the sentence in Turkish:


Looks good, and there were no problems while translating! However, I realize there is something wrong while working on the 85th page.


The problem is the difference between Turkish and English syntax, which the software cannot solve. Most translation memories show the first segment as “match: 100%,” and the translation software will skip it since it thinks the translation of that segment is complete, even though it is not.

The correct translation should be:


Trusting Translation Memories

If I had trusted the translation memory 100% as it told me to when it claimed a perfect match, I would have mistranslated the last example. After I translated the last sentence, however, most translation memories would have suggested two translations for the first segment: “If you use a translation software, you may realize that” with a warning like “match: 65%” from then on. I have to check each and every segment while working with translation software, even when it says “match: 100%.”

About Melıke Uzun

I’m a native Turkish translator living in Turkey. I have a degree in translation from Istanbul University. I have translated 6 novels, but primarily focus on technical, IT, medical, and legal texts.