Smokin'...!

Me, that is.

I finally got to bed a little after 1 am this morning and, thanks to Sasha the Canine Alarm Clock, was up again at 6:15 am. I've been pretty much hard at the medical item since then, and am taking the opportunity to note - while DejaVu exports the translation into a Word file - that I have finished the preliminary pass with a few minutes to spare. I now need to review the translation for spelling, etc., but that should not take long at all, seeing as how it's only about 6230 words (32,927 characters) long. :^)

Cheers...

It r-a-i-n-e-d today!

Real rain, even if it was for just about 5 minutes. Anyone who thinks this did anything more than knock the fire hazard level down by just a notch for the next few hours is crazy, but it was a jolly, honest-to-ghu rainstorm that plowed through here around 6 pm.

By that time, Drew, Shannon, and Huntur had left for Durango to go visit their friend Kevin, who had been driving the car when the accident happened. It's a good thing to have friends who are willing to drive 120 miles, round trip, just to come and lend moral support, if nothing else. Kevin apparently walked away from the crash with just a few bruises.

* * *Lee tells me she is down to her last shekels and has taken a job within walking distance for $6.50 an hour to keep body and soul together while she waits for something to turn up out of some phone calls and an interview or two.

* * *The phone calls asking me to do translations continue to pour in, and I have now put a lid on new work until I move what is currently on my plate off into the proper area.

The "standard" approach taken in a lot of fields (including translation) is to find someone to help with overflows, but pretty much most of the people I know who are good enough to let me get a good night's sleep while they do work I'll put my name on are busy, too.

Fellow LJer brenk puts her thumb directly on the dime, though, and it's something I've known for a long time: many people who claim to be translators aren't, uh, very good, though some of them have potential. The former group should be avoided at all costs, and anything you farm out to the latter group needs to be reviewed carefully until they are up to full speed.

The fact is, to build up a translation business the right way, I think you need to able to invest the time to get the rough spots smoothed out with the people who are freelancing to you, which is the case in the initial stages of any kind of contractor-subcontractor relationship. In my experience, that fact is at least doubly applicable for translation, and I don't know if I have the time to break in even one new person.

You may be right, I may be arrogant, I don't know.

Over the years, I've read a heck of a lot of translations written by other people. Back in my pre-translation days, when I worked at Plenum, there were freelancers who could be counted on to provide howlers on a regular basis. The initials of one such translator, MDF, became synonymous with translations so bad, you didn't have to be able to read or understand Russian to recognize them ("Hey, did you see that em-dee-effing paragraph in that article on tokamaks? Yeow!"). The usage came about quite naturally, since it was MDF who was the source of at least 50% of such idiocy.

Don't get me wrong: While some of the Plenum "stable" turned out swill, some others, such as AT (who did solid-state physics), DHP (who did technical physics), and RBR (who did astronomy) did very high quality work. However, I believe that in translation, as in other areas, Sturgeon's Law holds: 95% of it is crap.

Part of the proplem may be the nature of the beast: A lot of people in word-related industries have pretty swelled heads (and maybe that applies to creative industries in general), which makes dealing with them that much harder. I remember one fellow, V.R., who once came to me to complain bitterly about some edits I had made to a translation of his.

"You're wrong!" he exclaimed at the outset, and proceeded to explain - among other items - why the sentence "I was born at Moscow" was grammatically and stylistically correct, and why "I was born in Moscow" sucked dead bunny rabbits through a soda straw. Futhermore, the cosmonaut who was the subject of the translation was indisputably going to "fly toward Mir," just as it said in the Russian. My saying he was going to "fly to Mir" was, well... there simply was no way to adequately criticize it, said V.R., who by that time was glowing beet red from the injustice of it all.

V.R. is not at all unique, but you don't have to be a poor translator to have an attitude problem. I know, for example, someone who is a very good translator, and who works very fast, but don't ever try to make suggestions about his work: it's a waste of breath. Recently, when I gently pointed out that a particular concept was translated as "X" in a number of existing documents, he defended his use of "Y" in the one he'd just translated by saying he felt his rendering was superior. In this fellow's eyes, his "improved" rendering trumped the idea of consistent usage within a family of documents, and little you could say would have any effect.

Yet another part of the problem has to do, I think, with the fact that you don't need a high density of errors to make something look or sound like garbage. I was amused recently, after stumbling across a government web site for an initiative to build a real-time speech-recognition/interpretation device, to read that the first phase of the project was going to attempt to build something that would be accurate 90% of the time.

Since this hardware was billed as something you'd want to give to troops (or humanitarian aid workers) in foreign lands, I got to wondering just how happy I'd be if I was a grunt using one of these contraptions to ask a local about the presence of, say, enemy troop concentrations or mine fields in an area.

90% accurate you say? Haaaaaaa-ha-ha-ha-ha-ha-ha-ha-ha!!!!! "Is anyone in the village allergic to antibiotics?" The odds are only slightly better than Russian roulette.

Another way to look at it: OCR software that reads text with 99% accuracy can be expected to misread 12.5 letters per typewritten page (assuming 250 words per page and 5 characters per word). If you spread those errors out, that means 12 words or so out of the 250 on the page will contain spelling errors, which definitely does not represent "ready-for-prime-time" quality.

As usual, I've gone far afield and have almost forgotten the point I was trying to make, which I think has something to do with how tough it is to find people you can trust to do overflow work when you're up to your gunwales in assignments. There has got to be a way to manage it without giving in to the Dark Side.

Cheers...