If you grew up with holes in ya zapatos
You'd be celebrating the minute you was
havin' dough
The following message was posted this week to
an old item from a few months back
on the ProZ-TAUS debate, in which the Holy Trinity of L10N
Automation was convened to recite the by now stale message that translators
will not be replaced by computers (only degraded). It is an interesting
testimony because it is a counter to my thesis that post-editing will be
accompanied by lower absolute rates (i.e., both per word and per day, month,
and year). Check it:
Hi
guys,
I
realize this comment is really late, but speaking of PE payment, I just
finished a 2 year PE project that actually put some decent bucks in my coffers
and was so incredibly easy for someone with my experience and education in translation.
(10 yrs. experience, MA translation).
At
the volume I could do on it (paid at discounted word rate), I earned about 100
bucks per hour. However, I agreed to it only because of the major concession on
quality. This stuff was all internal and would not be published, so they didn't
care about the quality. They asked for "understandable" and
"good enough" quality.
"Good
enough" and "understandable" are subjective. I often found many
of the garbled, yet decipherable, sentences simply "good enough." ;-)
No
complaints from them.
However,
I have had agencies approach me with projects that need to be publishing
quality--and they want a discount for MT--my response is usually to look at the
MT version, find it unusable and needing of complete re-translation and offer
them 2 options:
1)
inflated word rate by 50% if they insist on use of the MT.
2)
regular rate if they want it translated from scratch.
Needless
to say, I don't normally get chosen for these kinds of ridiculous proposals.
;-)
MT
can be used well on the type of project I described, which I just finished
doing. It all depends on the content and the purpose of the target text.
I
personally think that MT has no place in publishing quality work--not yet at
least. No matter how many strides they make in MT, they are a loooooooooooong
way off from replacing humans.
Whether
PE becomes a viable market strategy will be up to us translators. I did it for
a bunch of easy, repetitive internal garbage that is of interest only to the C
suits of a particular company--and made very decent dough. (I also did not lose
my style--I can still translate at proper quality too.)
If it
becomes the standard for publishing quality work, well, I'll be looking for
another profession.
Sorry
for the anonymous post, I don't want this client to make a guess at who I am if
they happen to stumble on this.
Great
blog!
I don’t doubt this testimony is genuine.
However, I doubt that $100-per-hour of post-editing is in the cards for many
individual translators in the future. Two observations are in order:
1.- The market for translations is very
inefficient. I won’t go into the theoretical meaning of economic inefficiency,
but I will only point out a symptom of that inefficiency. Any translator can
confirm this: Rates offered vary insanely from one job offer to another, by
factors of 100% or more. Not even websites such as ProZ have managed to
constrain the spectrum of translation prices. I already hear some ninny raising
his hand to say: “Of course rates vary. That is true of any market. Rates will
vary according to language pair, experience, difficulty, specialization,
volume, regularity, etc., etc.” As usual, this is a truism that papers over the
fact that a variation of one or two orders of magnitude is way too large for a
rational market.
2.- In an inefficient market, there will be
huge outliers. Our anonymous poster is just such an outlier. My bet is that, in
the future, competition among cheap translation providers will tend to reduce
the occurrence of such outliers. This $100-per-hour rate is an inefficiency
that will be gradually scraped away as one or two or three low quality players
become slightly more dominant in their niche. (Incidentally, the fact that the
author of the message maintains his contribution anonymous confirms this. His
suspicion is that if his clients discovered that the same tweaking of MT output
can be obtained at a much, much lower price, they would turn to another
provider… because the service is a commodity.)
I agree with the anonymous author that
there is not a place for MT in material for publication… yet. There would be, perhaps,
if the MT people got serious about quality metrics. That way, when I came into
the office in the morning, with my Starbucks venti and munching on a croissant,
and read 57% accuracy, I would simply eject the refuse into hyperspace and greedily
keep the 95% matches. But since some MT specialists have claimed on this very
blog that any Google Translate match is equivalent to a 95% Trados match, you
will forgive me if I remain a skeptic.
I wish I were wrong. If I am, and I end up
as a post-editor making $100 an hour, I’ll be right there with you, oh
anonymous contributor, filling up jacuzzis with Cristal, raising the roof, and lighting
Cuban cigars with one-hundred-euro bills while Jay-Z’s “99 Problems” blares in
the background. I will gladly sell my soul, sans sales tax, if
any such there is to sell.
But I truly, truly doubt that this will
happen. The impetus behind cheap translation is precisely to lower quality
expectations and make the entire process as cheap as possible. You might think
that this is irrational, but the business model simply depends on volume, which
is why these companies are trying to convince Fortune 500 corporations to
translate Facebook status updates.
Miguel Llorens is a freelance financial translator based in Madrid who works from Spanish into English. He is specialized in equity research, economics, accounting, and investment strategy. He has worked as a translator for Goldman Sachs, the US Government's Open Source Center, and H.B.O. International, as well as many small-and-medium-sized brokerages and asset management companies operating in Spain. To contact him, visit his website and write to the address listed there. Feel free to join his LinkedIn network or to follow him on Twitter.
"Whether PE becomes a viable market strategy will be up to us translators."
ReplyDeleteIf this is the case, the chances are that translators will again eagerly cooperate with outfits pushing this business model, which is in my opinion insane and really destructive to our bottom line, just as they eagerly cooperated when they were asked to start using TMs so that they could be paid less for "fuzzies" and similar inventions out of while cloth.
I'm afraid your prediction is probably correct.
But I also think that all translation markets will not be affected equally.
In patent translation, the customers need translations that is as correct and precise as possible in order to prevail in court.
And you simply can't get good results from work that is based on editing of MT. Garbage in, garbage out. No matter what you do, the results of editing will be inferior to real translation that is based on talent, knowledge, experience and creativity.
Lots of observations. Let me see if I can state them briefly:
ReplyDelete1.- I don't think lower rates were driven by CAT. It may have been a factor, but not a big one. I understand why you make the connection, but I just think there was a lot of hype surrounding the technology.
2.- I think you *can* derive publication ready product via PE, but when you get to publication-worthy level, the productivity gain has been negated. It's the same thing you say, only arrived at by a different road.
3.- Incorporating it into your marketing strategy does sound iffy. I wonder whether "incorporating" it simply means you end up working for a large agency.
Great post, Miguel, and I am in clear agreement with many of the points you make.
ReplyDeleteOne point I feel is always left out the debate about post-editing, however, is that of not simply translation quality, but also of content quality. I am a translator, editor and ghostwriter who spent the first 20 years of a nearly 40-year communications career as a journalist so my translation and editing work goes way, way beyond simply transferring words in one language into words in another. My processes involve research, fact-checking, language usage and customs, etc., etc. And these are the qualities I usually get hired for, even when my rates are higher than those of my competitors.
That said, what I question about post-editing - I mean, aside from the moral and ethical issue of its subordinating Man to Machine (ever see "Terminator"?) - is the fact that there is no guarantee of content accuracy, even when the language is fully up to publication quality level (which doesn't often happen either, as your hundred-dollar-an-hour Mr./Ms. Anonymous implicitly admits).
What I mean is this: Widely used MT platforms are notorious for brainlessly misinterpreting the entire meanings of certain phrases, whole paragraphs or even entire documents. (Give Google Translate a try on a text you're familiar with if you want to get a quick and horrifying/hilarious example of this). And most post-editing fails to take this into account. PE's job is basically cosmetic: They hand you a text translated by a robot, and you simply fix up the language to make it palatable for native readers. In other words, your job is to keep the robot from sounding like "Me Tarzan, you Jane"...period. It is not to sit down with the original and see just how good a translator the robot really is from the standpoint of content meaning.
What this signifies is that the post-editor is hired to "beautify" an MT text that, from the standpoint of content quality compared with the original, might well be absolute nonsense. And if the PE job were expanded to include even the most summary check of original against translation, the whole economy of MT would be rendered ineffective and it would be cheaper and more practical just to go back to having real human beings do the translating in the first place.
Bottom line: The whole idea of the MT/PE duo is a travesty of professional translation.
Hello Dan:
ReplyDeleteI confess that mostly when I discuss PE I envision a scenario in which the P-editor has access to the original. But you're right, some of the scenarios envisioned by our gurus include setups in which the PE-er doesn't have access to the original and sometimes doesn't even actually speak the source language! In that scenario, the idea is that the accuracy is 95 or 96 or 97% ad all that is needed is a little syntactic tweaking provided by the target language speaker. I think that is insane, but I would then, wouldn't I?
The issue you raise of accuracy is, of course, also important. However, a little semnatics clouds up the issue. When PE advocates speak about quality, they actually mean accuracy and quality (understood as stylistic quality) all wrapped up together, or sometimes just accuracy. The definitions can get slippery. In any case, my view is that on both stylistic grace and accuracy, the technology still isn't there. I will try to keep the distinction in mi d in the future, since it is an important one.
Hi Miguel, nice work! One observation: $100 an hour, 8 hours a week, 48 weeks a year comes to $192,000 before tax. Hardly enough to smoke Cubans and fill up jacuzzis with champagne. And I think that really top notch translators make close to that with or without PE.
ReplyDeleteOh, I think 192k a year is enough for Cuban cigars (I've had enough of those with lower incomes). It is not enough for filling jacuzzis with Cristal, although why anyone would do that is another question. Furthermore, how one could read the statement as anything but sarcastic is also a worthwhile matter for speculation.
ReplyDeleteAbout money I never joke.
ReplyDeleteMiguel, I'm ill at ease with (1) the anonymity business and (2) your anon poster's claim that "(I also did not lose my style--I can still translate at proper quality too.)" Not to mention (3) the cynicism. I'm no starry-eyed pollyanna person, but even so would worry about that terrible wasting disease — the cynical attitude embodied in "we know it's crap, but hey, those idiots think it's good enough").
ReplyDeleteMy experience of good (direct) clients who pay well (generally substantially more than $100/hour) is that they take their texts very seriously and get turned off fast by cynics.
Chris
Hello, Chris:
ReplyDeleteThanks for visiting and posting your comment. Yes, I agree with you on all three points. I've been wrestling with the anonymity issue for a while now, and I still haven't arrived at a satisfactory answer, although the growing number of messages to this blog will probably force my hand one way or another pretty soon. Point taken regarding cynicism. Perhaps my own response was also a little flippant. This was an indulgence on my part because the poster implicitly was making a lot of points I tend to make. Still, thank you for keeping us honest.