Many aspects that are obvious to us as language service providers are not necessarily clear to our clients. We are frequently asked why there are two separate systems for storing translations: the translation memory system and the terminological database.
What is the difference?
Both systems serve distinct purposes, and it is most effective to use them together:
A translation memory system stores translations and supports translators by providing text segments that have already been translated and can be reused. To enable this process, each translated text is stored in the translation memory (TM) segment by segment. Before translation, new texts are divided into individual segments so they can be compared with the translation memory. A segment typically corresponds to a sentence, as punctuation helps define where a segment begins and ends.
During translation, the translation memory system uses an algorithm to search for matches in the translation memory for each new segment and displays relevant results. Translators can then insert these suggestions and adapt them if necessary.
In a terminological database, only individual words and very short phrases are stored. This database functions as a dictionary and is used to record targeted information such as technical terms and definitions. You can also store additional metadata, such as “prohibited terms,” contextual information, or even images.
Terminology is first defined in the source language and can then be applied in the various target languages. The goal is to maintain the greatest possible control over the key terms used in a text – and, as a result, ensure consistency. Consistent terminology is an important factor for clear corporate language. It helps prevent misunderstandings and ensures that texts are easy to understand. By using consistent and clear terminology, companies shape their brand and set themselves apart from the competition.
The question remains: why not store individual key terms and short phrases, as well as longer segments, in a single system?
Why are there two systems for key terms? Would one not be better?
A TM focuses on quantity, while the terminological database focuses on quality. The TM ensures that you do not have to retranslate or retype existing texts repeatedly.* Terminology – meaning each individual key term within a text – is a qualitative specification designed to prevent misunderstandings about the intended meaning. Accordingly, the two systems also differ in their structure.
Translators now combine both systems: On one hand, you receive translation suggestions from the TM, which you can adapt as needed. On the other hand, the terminological database linked to the system notifies you when a defined term appears in the segment to be translated. You do not need to research this term, as it is displayed to you immediately when it appears in the text. A reliable translation memory system also enables automatic verification of the terminology used against the terminological database as part of the quality assurance process.
* The suggestions are based solely on character similarity. They do not indicate whether the content is actually correct in the context or which variant should be preferred if there are several different matches.