Insight into how to maintain TMs for languages I don't speak.

We are looking into introducing Trados to our organization and we would like to maintain memories for a number of languages that no one in our staff speaks. Our Spanish-speaking staff can create, update and align memories, for example. This way we can send translators memories from previous jobs when a new job comes up. We would like to do this for other languages. But we don't know if there is any way to do so for language that we don't speak?

We would appreciate any insight anyone can provide.


  • Hello

    Indeed these are very good questions.
    I used to be a translation project manager working with Trados Studio every day. I was also tasked to create and maintain memories with languages I don't speak, and align documents. It is always tricky a bit but manageable, in my experience. 

    As for the alignment, my suggestions:
    - You can align a documents even if you don't speak the languages, although it's always safer if the person doing the alignment speaks at least one of the languages.
    - The tool aligns the documents based on their structure, i.e. the first segment of the source will be paired up with the first segment of the target. Therefore the more similar the two documents are in layout and structure, the better the automatic alignment will be. With this in mind, it is usually a good idea to check the documents against each other before the alignment, and modify their structure to make them more similar.
    - If ever you are suspicious of a pairing, the best is to contact a translator or if not possible, use machine translation to see if the two sides of the pair match.
    - If you are not fully sure of the quality of the alignment, don't be afraid to apply higher penalty than the default 1%.
    - It's best to align Word and Excel documents, and you can also compare them easily side-by-side outside of Studio, if needed.
    - Always save the alignment as you go along, and save it when ready as well. Then you can load it into a new memory with some penalty, and you can start working with it, but in the meantime if you want to be extra safe you can send the alignment project out to a translator to review it, and then load it into a new TM, once it is verified.

    Translation memory maintenance:
    - Always be consistent. If you have memories in different languages for the same subject matter, treat it as a group of memories.
    - Always update them together, so that the content remains similar. If you have to correct something or delete a few strings, check these issues in all the memories of the group, and make the necessary changes everywhere.
    - Here as well, if you need to find, delete certain strings, ask for help from a native speaker, or if not available, use machine translation to get the gist. For textual changes however, always rely on a translator, check it with them.

    Hope this helps.
    Let me know if there are more questions or concerns.

    Best wishes,

    Greta Magyar - Customer Success Consultant - SDL

Reply Children
No Data