Logo BTS Empresa de Tradução

Automatic Translation vs. Computer Aided Translation (CAT) Software

Sobre o autor

BTS: Luiz Fernando Casanova Doin

Luiz Fernando Casanova Doin

é fundador e sócio-diretor da BTS – Business Translation Services, empresa de tradução sediada em São Paulo com mais de 15 anos de existência e mais de 5.000 clientes atendidos. Formado em Administração de Empresas pela Universidade de São Paulo (USP) e em Comunicação Social com ênfase em Marketing pela Escola Superior de Propaganda e Marketing (ESPM), é um apaixonado pela área de tradução e um estudioso do setor que procura aliar seus conhecimentos acadêmicos e experiência à prática tradutória.


One recurring question from our clients is, “Do you employ automatic translation?” Most likely they are thinking of tools like Google Translate, Babelfish, and alikes — free online services where all one has to do is to enter their text in one language, and have it almost instantly translated into another. No, we don’t use such tools simply because the translation quality is bad. Unfortunately (or fortunately for us, translation companies) there are no miracles. These automatic translation contrivances serve only to provide a general idea on the text being translated, however they are far from offering professional translation.

On the other hand, we use the so-called CAT (Computer Aided Translation) Tools (Trados, WordFast, Déjà Vu, etc.), that are programs that help (a lot) translation being done using a computer. Differently from automatic translation, such programs don’t translate anything on their own. Translations are done by live “bone and flesh” translators who use such software as their translation platform.

Yet, if that Computer Aided Translation (CAT) tools don’t translate on their own, what’s the point in using them? Basically to increase productivity and to ensure terminology consistency. Upon translating with the aid of such software, as we translate documents for a specific client, it builds a database with the translations we’ve done so far. The more we translate, the larger this database will grow. We, agencies, call these databases TMs – Translation Memories. They are actually files that contain all translations done, divided into smaller segments (usually one sentence each), containing the original text and how we translated it.

What are the advantages of having a translation memory for any specific client? Many… I’ll mention first the financial advantage: the CAT Tool software queries the translation memory; if it finds identical or similar phrases in the database, it will suggest a translation. So the program identifies segments that have already been translated, and automatically proposes the translations used then. In repetitive texts, productivity gains are staggering. The translation agency saves a considerable time, avoids a lot of unnecessary rework, and can transfer that benefit to the client as a discount on repeated texts (which we call “matches”).

I believe that the savings for both the translation company and the client is obvious. However that’s not all of it. There is also the extremely important issue of keeping a constant terminology. I’ll show with a practical example: we work for some audit firms that have their own “standard” translations for the balance sheet items. For instance, Company X translates “cash and cash equivalents” into Portuguese as “disponibilidades”, while Company Y insists that the proper translation is “caixa e equivalentes de caixa”. For us there is no problem, as we keep a separate translation memory for each client. If by accident a less alert translator uses “disponibilidades” as the translation of “cash and cash equivalents” for Company Y, the software will show them that the translation memory indicates the proper translation as “caixa e equivalentes de caixa”.

When one project involves several translators, CAT Tools are the only way to ensure that all of them will translate a specific term exactly the same way. Otherwise we might have Translator A using “disponibilidades”, Translator B preferring “caixa e equivalentes de caixa”, Translator C choosing “caixa e bancos”, and finally Translator D opting for “caixa e equivalentes a caixa”. In other words, four translators, four different translations. Imagine what it would take to manually consolidate this terminology, without the proper software. Wouldn’t it be insane?

Compartilhe esse Post:

Share on facebook
Share on linkedin
Share on twitter
Share on pinterest
Share on email
Share on whatsapp