How Effective Are Proofreading Methods Currently Practiced by Translation Agencies?

Every few months or so I receive an offer to proofread a translation of a patent or another document from Japanese, German, French, or Czech to English from an agency that found my name in the American Translators Association (ATA) directory. I still accept translation offers from agencies, but I don’t proofread translations that were done by other translators for translation agencies anymore. It is just too much of a hassle for me. For one thing, I have no idea whether the translation will be very good, pedestrian, passable, horrible …… And you can’t really charge much for this kind of work because proofreading is generally paid quite poorly. So why bother.

I think that translators who still accept this type of work are mostly beginners, or occasionally perhaps also more experienced translators who are going through a slow period. Incidentally, the last time when I was still accepting proofreading assignments from agencies, for 30 dollars an hour, was when Ronald Reagan was the president and the word Internet would be completely incomprehensible to most people.

I think that there are probably some translation agencies out there that really know what they are doing and have a well designed workflow, which would include also a well designed proofreading method. These would be translation agencies, mostly very small ones, that specialize only in certain languages and certain subjects. But from what I have seen, I think that these agencies are in a minority. Most translation agencies that “specialize” in every language (including every language direction) and every subject, which would be the majority of them, use the following method.

Let’s say that an agency receives a document from a law firm in a “language of limited diffusion” such as Czech, Norwegian, or Croatian. Most of the time, the person handling the translation will have no idea what is in the document. It could be a patent application, but it could be also for example a new law.

One option a translation coordinator has in such a situation is to find a translator for this language online. Relatively few translators have their own website, but many have for instance “a page” on Proz or on Gotranslators. Many translators based in US are listed in the ATA database. I have a listing there too and I usually receive one or more offers of work from this listing a week (although this will usually result in only a few jobs a year). Some coordinators send the same e-mail about a prospective job to several translators so that whoever answers first ( a “first responder”) and has an acceptable rate gets the translating job, and another translator (a “second responder?”) is offered the proofreading job.

Since the agency really has no idea whether the translator who was picked from a database and who got the translating gig is any good, proofreading by another translator seems like good insurance in such a case. The problem is, the agency has no idea whether the person who accepted the proofreading gig is any good either.

Let’s assume that the first translator was very good. There was nothing to correct in her translation, not even one typo or omission. She’s a regular genius. And let’s say the proofreader is not too bad either, but not as good as the original translator. What is he supposed to say? “I did not change anything because the translation was excellent, I could not even find a single typo. Here is my invoice for 150 dollars for 3 hours of proofreading.”

Is he going to get paid now “for nothing”? Maybe, but why take the risk. The proofreader will probably try to change something “to earn his keep”. A smart proofreader would make a few unimportant changes here or there without causing too much damage to an excellent translation to make sure that he will get paid. But some damage is still possible, or even likely. A not so smart proofreader could mess up the translation considerably. A really dumb translator will turn a good translation into a FUBAR production. Since the agency coordinator does not understand the language of the original document, she has no idea whether what the proofreader is saying is correct or not. So sometime they try to run the proofread translation by the original translator one more time and the original translator is asked to “accept” or “reject” the changes. The original translator is expected to do this for free of course. It happened to me a few times and every time it made my blood boil (I am easily excitable when somebody tries to change something in my translation. It’s one of my many personal flaws).

“Don’t you want to know what the proofreader thought about your translation?” said an agency owner to me once. No, I don’t. I don’t know the proofreader from Adam. The chances are he is a total nincompoop if he has to work for the low rates that agencies are willing to pay for proofreading.

If we try several combinations of translators and proofreaders who are at a different skill level, the following equation should also make sense: a poor translator + a good proofreader =  a much improved translation. This could certainly happen, but given that most experienced translators don’t like to waste their time on low paid proofreading gigs, it is not a very likely scenario, although not an impossible one.

I think that a much better method is a method that is based on real knowledge and experience instead of the all too common “first responder” + “second responder” concept. A coordinator from a small, specialized agency (who is usually also an experienced translator) is much more likely to find a good translator for a given project, even in a language that this coordinator does not know. Let’s face it, the kids who work for large agencies as coordinators don’t really know anything about anything, and they will be gone soon anyway once they learn something, either running their own agency or doing something completely different.

A person who happens to be an experienced translator will proofread a translation that was done by another translator very carefully, but there is no pressure on him to change anything at all if it is a really good translation. And he will only make changes that need to be made.

Because he knows that once you have found the right kind of translator for the right kind of job, although you still have to proofread it slowly and carefully looking for typos and omissions, if it was a good translation, you don’t need to change anything at all.

The best proofreading method is a method that is based on finding a really good translator first. How likely is it that a “first responder” who charges a “very competitive” rate (lower than what an experienced translator would charge) is a really good translator?