CAT tools: what are they and how do I use them?


Firstly, it’s important to distinguish between machine translation (MT) and computer-assisted translation (CAT). Machine translation is when you enter a text, or an entire document, into an application such as Google Translate, and receive the translation of that text, performed entirely by a computer, with no human involvement. While MT is constantly being improved, it has serious limitations in terms of quality, accuracy and other more intangible considerations such as “flow” and “flare”. If you are interested, DOUGLAS HOFSTADTER published a fascinating piece in the Atlantic, exposing the“Shallowness of Google Translate”.  I’ll just leave you this quote from that excellent article as food for thought: “There’s no fundamental reason that machines might not someday succeed smashingly in translating jokes, puns, screenplays, novels, poems, and, of course, essays like this one. But all that will come about only when machines are as filled with ideas, emotions, and experiences as human beings are. And that’s not around the corner. Indeed, I believe it is still extremely far away.”

Computer-Assisted Translation is an entirely different technology. A computer-assisted translation is performed by a human translator (ideally an experienced professional), who uses various applications to help her to work more quickly, accurately and consistently. The most complex of these tools are translation memory applications (essentially databases in multiple languages) and terminology databases (which enable translators to locate the correct terminology for the relevant field and ensure consistency within a document and with past and future texts). There are also text aligners, a handy tool for viewing the source and translated texts side-by-side for review purposes. 


When you commission me to work on a document, I will use a CAT tool (such as SDL Studio or Swordfish by Maxprograms) to create a translation memory database and a terminology database with your company’s name. As I translate, the text will “populate” these databases. Once I have reviewed the translation, and made any changes or corrections, the database will be purged of any unwanted terminology. If you request any changes to a translation, those changes will also be entered into the databases, for future reference. The next time you hand me a text for translation, I will apply the databases before starting. If there is any overlapping terminology, the databases will find it. Sometimes, there might be entire strings of previously translated text. Sometimes, there might be what are known as “fuzzy matches”, where there is no exact match, but there are similar sentences in the database that are reusable once tweaked accordingly. Generally, I will apply different rates to the different match (from “no match” or completely new text, to 100% match, where text is identical to that of a previously translated document*. You will save money in the long run, especially if I can leverage previously translated texts to harvest repeated content. Evidently, this will apply more to certain types of translation than others (technical texts or periodic reports that consistently use similar content are likely to benefit more than literary or artistic writing). Feel free to ask me about my rates. Before you approve a project, I will always provide you with a quote, based on word count (excluding any previously translated text I may have in the database), deadline and complexity.


*It is my practice to only apply discounts for the use of CAT tools when the material I am harvesting to populate databases was either translated by me or approved by me before starting the new project. I will not enter material that has previously been poorly translated or is full of typos or errors into the database, so I won’t be able to use it.


Swordfish Translation Editor ®