Computer-Assisted Translation software, commonly known as CAT tools, are extremely useful resources which help translators in the translation process.
These are not by any means the same as machine translation engines. Neither do they replicate or replace human’s unique ability to interpret meaning. They «simply» compare, store and retrieve data already translated by the user – i.e. a human translator.
So how do they work?
At the start of a translation project, the translator uploads a document or batch of documents into a CAT tool, which split the original document into segments (either lines, sentences or paragraphs). The interface then produces a two-column table of individual segments, with the source in one column and an empty space waiting for the translation. These tools filter out formatting and other tags, presenting only the translatable text, protecting the tags and code and allowing the translator to export the translated text with the original formatting intact.
Amongst the most critical functions of a CAT tool are the databases of prior translations: termbases (TB) and translation memories (TM). Once the translator has finished the translation, these databases are updated to serve future projects. They both store data from previous translations, though they have two different approaches.
Termbases
Once the translation has been done, the translator manually selects relevant terms and their translations from the project. These terms, which are usually specific to the client or to the client’s industry or field of specialisation, can be used in potential future projects for the same client, ensuring consistency and brand sense. This is what we call a termbase (‘terminology database’).
Translation memories
On the other hand, a translation memory stores and classifies each individual segment and its equivalent translated segment (as a reminder, a segment is a whole line or sentence, normally). In short, it saves the whole text, only split up. Translation memories are very useful to maintain consistency in future projects that share some segments or structure.
However, they need to be used wisely, as context will likely need attention to detail and generally it’s not a good idea to blindly accept all matches in the memory.
When the translator is working on the project, the computer-assisted tool analyses the new text, looks up for matches and «reminds» them of previous translations that are stored in the databases. The translator can then check those matches, verify the context and -if everything is looking good- choose the same translation strategy, use the same tone or keep the structure used in previous projects.
Hence, these software tools allow the translator to never forget even the slightest previous work (like an elephant, that never forgets!). These are a superb solution to keep consistency and style throughout all projects from a specific client.
The benefits
As we have just seen, CAT tools have nothing to do with automated or machine translation. It all comes from the translator’s own work and expertise.
They can be extremely useful resources for both the translator and the client, as they allow consistency and accuracy in terminology and style, which are among the most critical aspects to deliver a flawless translation.
Leveraging the benefits of CAT tools in your projects is always advisable, but it is crucial if you have a regular flow of content that needs translation or if you’re translating technical, software or other specialised content. Shall we create some reference files for your flows?