I still remember the first translation I ever produced for a paying client. Back in 1987, I was given a 19-page technical document to be translated into English for a major electronics company. I think I had a Commodore 64 or Amiga at the time, but it was no good for typing up a professional translation. So, I used my Brother typewriter.
It was a lengthy process – and torture compared to how easy it is today to edit documents in MS WORD. Whenever I made changes during the proofreading process, I had to retype the entire page in question, and often the following page as well if the text had become longer and spilled over to the next page.
Looking back on those primitive means in 1987, I am thankful for PCs and Microsoft, as well as any other key maker of hardware and software that have made life, and work, so much easier. But with every silver lining comes a cloud, and in the translation business that cloud is known as CAT, Computer-Aided Translation.
CAT comes in many different forms, that is programs: Trados, wordfast, DVX, SDLX, Across, etc. It is not a form of machine translation, which doesn’t work and never will, but a management system that accelerates the actual – human – translation process. It stores previous translations and creates a rich database of terms, phrases and entire sentences or paragraphs that the CAT program will bring up automatically in your current translation project if it detects a suitable match.
This way, the translator using CAT ensures consistent style and terminology when working for the same client over a period of time. It also helps the translator to save time: a perfect (100%) match, for example, means that the entire segment (phrase, sentence or paragraph) found in the database (“TM” or Translation Memory) can be incorporated into the current translation, saving the translator the trouble of having to think about and write the translation of the segment all over again. Even a less-than-perfect match can still be a time-saver, as, perhaps, only a single word may have to be changed to make it a 100% match.
I have to admit that I was extremely skeptical at first. I refused to buy and install any of those CAT programs, because my own “human TM” would always work best. Actually, it still does, because I always find that my own brain detects potential matches in a new document even before the TM spits it out. I have a very good memory, but I would still have to retype the segment regardless. CAT, at least, relieves me of the need to retype it all over again.
Over the years, I have tested and worked with all the major CAT tools available out there: Trados, DVX, SDLX, wordfast and so on. DVX has always been, and continues to be, my all-time favourite. It is the most efficient and most reliable CAT program in the marketplace. Trados, in all its various incarnations, has always been beset with problems and glitches, which would freeze up or crash your computer on numerous occasions. From the daily cries for help that Trados users post online, I know that these problems persist to this day. But to accommodate clients who insist on the Trados format, I use wordfast, which, oddly enough, never causes any of those problems even though its approach and coding are virtually identical to Trados. (It also costs only a fraction of what Trados sells for.)
But despite the much-appreciated assistance we translators get from CAT tools, CAT has been instrumental in what I call the Balkanization of the translation industry. Let me explain what I mean by that.
Translation agencies, and some direct clients, will insist that the translator use a specific CAT program to the exclusion of all others. This can be quite the problem for the translator. When a new client like that comes calling and makes such demands, it is impossible for the translator to acquire the CAT program in question on such short notice, which means that he or she will lose this potential client right away. Most major CAT tools claim some level of compatibility with formats generated by competing products (e.g., working on a Trados format in DVX), but it is never a perfect solution.
From the client’s perspective, it may also be a loss. The translator may have been the best one for the subject matter and type of text; going with another translator, simply because the other one happens to have the “right”, and fully compatible, software, may force the client into a relationship with a translator who may not be as professional or qualified as the first one. In the worst-case scenario, the client is handed a translation that is absolute rubbish – and using the required or desired CAT tool won’t change that. A CAT tool, after all, is only as good as the translator using it; if the translator is a “wannabe translator” without any training or actual talent, who cannot string words together to form even a single sentence without making one mistake after another, the “translation” cannot be turned into a professional document even if it is run through the most expensive CAT software.
So, what we have today is a Balkanized industry. There are several formats and programs, and it is becoming ever more difficult to match translators and agencies/clients as a result. The other day I saw a job ad by a translation agency looking for new translators to be added to its database, but only those who use a fairly unknown program, Across. It is actually not a bad program, but compared to Trados and DVX, and even wordfast, Across is a bit player at best. Meanwhile, the agency in this case has limited the size of the pool in which to fish for new talent. This is not only short-sighted and dumb, but also unprofessional and damaging to the entire translation industry.
Apart from the agency depriving itself of good translators, it is also translators who suffer the consequences of its actions, because they will not be able to share in the work that the agency may have to distribute over the coming weeks, months or years.
What is more, a growing number of agencies recruit translators on the sole basis of the CAT tool they own and use. CAT ownership or proficiency is not a professional qualification. A degree in translation or accreditation is – or actual talent, for that matter.
I am not saying that owning a CAT tool makes you a lousy translator. What I am saying is that when you select translators on that criterion alone, you are more likely to end up with “wannabe translators” than with top-notch talent.
These fissures created by the various “CAT communities” in the industry are not healthy – not for clients and agencies, and certainly not for translators.
Interesting post. But what about the scaling down of rates paid on the basis of degrees of matches? Do you accept it?
I, on my part, am always against it. However, my stand is academic, as I do not possess any CAT tool nor do I intend getting one. I am happy without it.
A translator invests good money to procure a CAT tools and loses money in the bargain?
Regards,
N. Raghavan
Posted by: N. Raghavan | January 16, 2010 at 06:38 PM
I used to be totally against "TM discounts", but have granted them to some -- reasonable -- degree in the past (only 100% perfect matches at best, but no discounts for anything under 100%).
Most CAT tools are extremely expensive, which is why translators should really be careful about granting self-defeating discounts for TM matches (some greedy agencies insist on discounts even for less-than-50-percent matches, which is absolutely ridiculous).
Posted by: Werner Patels, M.A. | January 17, 2010 at 12:24 AM