About me
Over more than twenty years as a freelancer and a few extra years as an in-house translator with clients and agencies, I have translated and edited more than twenty million words in the fields of oil & gas, energy, legal, and IP from Spanish, Portuguese, French, and German into English. Originally from the backwoods of Salzburg, Austria, I now reside in Querétaro, Mexico, after many years in San Francisco, California and a short stint in Rio de Janeiro, Brazil. Whenever I am not translating, I like to take off to explore the more remote corners of the world.
Albert Rieger – Allingua Expert Translations & Language Consulting
Spanish – Portuguese – French – German
Querétaro, México
Have CAT tools had an impact on the overall pay of translators? Certainly, in case one blindly accepted the oversimplified rationale of the new agency model of full vs. fuzzy matches. Did they impact rates per se? I don’t think so.
At the same time, I am quite certain that computer-assisted translation tools did have a great impact on the earning potential of translators because CAT tools are just so symptomatic of what Steve routinely calls the “new agency model” or what I refer to as the “industrialization” of the business of translation.
I still do remember the times when translation agencies/firms were usually run by people who were either translators themselves, or non-translators who still had some foreign language background, or at the very least people who had more than just a superficial appreciation for what translators actually did or what translation actually involved. In other words, people who were aware of the “qualitative” aspects of translation.
Those were the days when I first moved to the States, and as a German speaker, what immediately struck me as sort of odd was the fact that the term Branche, as in Übersetzungsbranche, would be rendered as “industry” in English, as in “translation industry”, even back then. A business, yes, I thought by myself, but there was clearly nothing industrial about the activity, in my mind. This was nothing like car manufacturing or steel production, I mused. But hey…
But if the translation industry wasn’t an “industry” per se back then, it did not take a long time to become one in the truest sense of the word. Now, it’s an entirely different ballgame. By now, the “translation industry”, at least the mega-agency part of it, is run by people who don’t seem to have the slightest clue about translation, about translators, to say nothing of having any background in foreign languages. Needless to say, they would not know a thing about the “quality” of translation, and truth be told, they could not care less. Seems like this new corporate CAT/MT universe is entirely populated by business people, software engineers, mathematicians, and whoever not… anything but language professionals or, God forbid, actual translators.
And while there weren’t any industrialists in the good old days, there sure is no shortage right now.
And what these self-anointed translation industry leaders all do care about are the “quantitative” aspects of translation, which is really the only relevant aspect when it comes to “selling high and buying low”. With this mindset, it is really the only thing they can care about, even if they really wanted to care about anything else. The pressure of shareholders certainly doesn’t help, either.
So, for the quantity aficionados, things are actually very simple: a text is nothing more than a quantity of “translation segments”, which can be stored and then recycled as necessary. For them, the fact that a translation segment is already in memory simply means that it doesn’t have to be translated anymore, it just needs to be inserted, and voilà, that’s not translation, right, so why should the translator be paid a second time? And if something similar has been translated, well, according to this logic, that’s a “fuzzy match”, which should not be paid in full, because, in their non-translator world, it does not require the same translation effort as a full match. Simple as that.
When I am talking about the loss of quality, I am not only just talking about the aspect of whether a translation is good or bad. I am also referring to aspects which greatly affect how long it takes a translator to properly translate something… like extensive formatting, readability of the text, or translatable text interspersed with non-translatable English text which needs to be reproduced, anyway, or pre-translated elements which may or may not be correct in a new context. For the new breed of agencies, unless they are “translatable units”, none of these qualities or elements seem to matter… at least not when it comes to payment to the translator who has to deal with them.
For the translation industrialists, there are also no longer any words with shades of meaning and register, just “units”. Just like there are no translators, just “vendors”. I have yet to be actually called a “translation unit vendor” by any of them, but I know the time can’t be far off.
And with their extremely near-sighted view (if one can even call it “view”) of what translation actually is – for them, this simply means an assembly-line production of translation units, which unfortunately still requires translators as some sort of at best semi-skilled labor, hopefully to be rendered obsolete in the very near future – it should not really comes as a surprise to any of us that they have been pushing very hard to impose this concept of theirs, not only as the best one, but as the only viable one, really.
Nor should it come as a surprise that the ones they have been able to convince of this outlandish statement are mostly newbie translators who simply do not know any better, who believe that “CAT tools are the norm now”, which is a myth the new agency model is all too happy to perpetuate ad nauseam, despite quite a bit of evidence to the contrary.
Whenever agencies ask me why I don’t use CAT tools, I personally always respond the following, in this order:
- Stability issues with CAT tools: I am not looking forward to unexpected software conflicts and computer crashes with CAT tools, which have been very well documented, so by not using them, I avoid the risk of being unable to complete a project on time, which is paramount for me.
- Based on many years of experience and actual try-outs of different CAT tools, they offer no time savings for me for the kinds of projects I usually work on.
- Through “segmentation”, they interfere with my personal “creative flow” when translating and actually decrease both my translation output and quality.
All of these caveats actually do reflect my real-life experience with CAT tools, so it’s not like I did not try. And it’s not like I am making things up.
Curiously and fortunately, after mentioning these shortfalls to agencies, I have never had any agency probe any deeper into any additional reasons and motivations I might have, so I never had to mention the other serious concerns I have about CAT tools. It is probably a good thing that they don’t really know what projects CAT tools are actually helpful with and for which ones they are a waste of time (my point 2) and what the creative process of a living and breathing translator actually entails (my point 3). Something like that would require some knowledge of how a translator ticks, and that is simply not something which characterizes the new agencies.
In summary: CAT tools may have directly cut down on translator’s earnings in some cases and not in others. But that’s not the whole story, and certainly not the most significant one.
I do believe that, indirectly, but much more significantly, CAT tools have certainly had an adverse impact on translator pay because, rather than being used to enhance the productivity, consistency, and thus the income of translators, which is how they were initially marketed to translators, they were quickly hijacked by the emerging corporate LSP model to line the pockets of a new breed of translation business operators (almost exclusively non-translators) and their shareholders for which only “quantitative” aspects of translation exist to the exclusion of any “qualitative” issues.
And while translators were and are still debating the pros and cons of CAT tools under different scenarios and circumstances, CAT tools immediately made perfect sense to the new operators: segments which had already been translated did not need to be retranslated, regardless of how inaccurate or outright wrong they were in any given context, period. More crucially, they did not need to be paid again, under any circumstances, and similar translations did at least not have be paid in full. And as long as this new business model could be strictly imposed on translators, or most of them, there was almost unlimited potential for increasing profits… those of agencies, mind you, not those of translators.
Keeping this in mind, I firmly believe that CAT tools were the actual killer application for the emergence of the new agency model, just like Lotus 1.2.3. and WordStar were the killer apps for the IBM PC. And in the process, everything else that was incompatible with this approach was stripped from the translation process, and the art of translation was downsized to a mere production of “translation units”, something that could simply be measured and then hopefully be multiplied, at best ad infinitum and into ever greater riches.
Unfortunately, as “killer” as CAT tools might have seemed in the beginning to translators, especially the way they were presented to them by CAT tool vendors, I can see that it is just now that many translators start realizing who is making a killing… and who is getting killed.
And that is something that has impacted all of us, whether we are actually using CAT tools or not.
Patenttranslator's Blog 
Interesting reasons:
“Stability issues with CAT tools: I am not looking forward to unexpected software conflicts and computer crashes with CAT tools, which have been very well documented, so by not using them, I avoid the risk of being unable to complete a project on time, which is paramount for me.”
Well, I can’t remember the last time I had a stability issue/software conflict with DejaVu, but there you go.
“Through “segmentation”, they interfere with my personal “creative flow” when translating and actually decrease both my translation output and quality.”
A valid point, if you’re receiving segmented texts from agencies. However, if you’re doing your own segmenting, it’s possible to segment according to paragraph rather than sentence, which gives you more leeway to be creative – at least, that’s the case in DejaVu, I don’t know about other software.
LikeLiked by 1 person
By: Alison Penfold on June 21, 2017
at 9:43 am
This comment for some reason did not make it past the WordPress software onto my blog, so I am posting it from my email manually.
But it is not my comment, the comment was written by Kenneth Kronenberg:
Thanks to Albert Rieger for this very interesting post — and to Steve for posting it. I think he is absolutely on the mark when he writes that “CAT tools were the actual killer application for the emergence of the new agency model” for exactly the reasons he states. Given the battle over who owns TMs, most translators have at least a tacit understanding of what those reasons are. One of the things that is really scandalous to my mind is how university translation programs are conditioning students to enter an industrialized market saturated with this mindset. And of course, the developers of CAT tools are more than happy to subsidize this miseducation.
Those who have read my paper “Translation in the Twenty-First Century” (http://www.kfkronenberg.com/Translation-in-the-Twenty-First-Century.pdf) know that Big Translation has been a major concern of mine. Wherever we look, the technology that was meant to make translators’ lives easier has actually fostered the consolidation of Big Translation — at the expense of actual translators. But this is a much larger issue than I want to pursue in this reply.
However, I do think we need to start asking some fundamental questions: Is there in fact a cadre of rising translators at this time? If so, what is their conception of what translation is? How long do we think that they will be satisfied doing the kind of piecemeal segmental work that Albert describes, not to mention post-editing? How many young people just out of training do we think will eventually make it into the so-called “premium market”? How many will fail, for whatever reason? And to the extent that young translators learn to think of translation in terms of units, as Albert suggests, will they not be more apt to go into translation management? If we are doing mentoring, what world are we mentoring them into?
I’ll leave with one parting shot from a recent Slator article about RWS Group:
In its earnings comment, RWS highlighted the continued growth in its core patent translation and filing business. International patent filing was traditionally done by patent attorneys who engaged freelance linguists to provide translations. Over the past decade and a half, RWS streamlined that process, creating a market it now largely dominates.
More big talk from an industry giant? Probably. But the intention is clearly there, and those who now tout the premium market of private clients may be irresponsibly setting up the next generation for a fall.
Best,
Ken
LikeLike
By: patenttranslator on June 22, 2017
at 7:10 pm
“International patent filing was traditionally done by patent attorneys who engaged freelance linguists to provide translations. Over the past decade and a half, RWS streamlined that process, creating a market it now largely dominates.”
Ha, ha, ha. Transparent corporate propaganda at its worst.
This freelance linguist has been providing patent translations directly to patent law firms for 30 years …. and I have never been busier than I am now.
The market for patent translations, both for prior art research and for filing, is so huge that no single translation agency can possibly claim to “dominate it”.
I should add that I have known about RWS since the late nineteen eighties, but I could never work for them even 30 years ago when I was a total beginner working only for translation agencies because even then, their rates were much too low for me.
I wonder what kind of translators they use.
LikeLiked by 1 person
By: patenttranslator on June 22, 2017
at 7:19 pm
I would like to make 3 serious points here:
1. In the early 2000s, when non-translators like SDL’s owners entered the translation market (unknown to them until then!) with the “fuzzy match rebate” concept, they MIGHT have been sincere.
But, as they did not belong to the translation world, they did not know that AGENCY RATES ARE ALREADY A HUGE 50% REBATE OVER THE “BASE RATE”.
Translators not knowing how to draft a proper invoice (and often being trained by our very exploiters, i.e. agencies), they usually FAILED TO INDICATE that THEIR OWN ‘BASE RATE’ was the END-CUSTOMER RATE.
Junior translators usually TOTALLY IGNORE what the average end-customer rate is when they start working, the only “rate” given to them, either by colleagues or by agencies, being the AGENCY RATE, which is ALREADY A REBATE – and a too huge one, because 50% gives the intermediary the idea that he is our EMPLOYER (see point 2 here below).
So any freelance translator’s INVOICE should read like this:
—
a. Base rate (i.e. end-customer’s rate – we should say ‘fee’ actually): something like 0.20 EUR/USD per source word
b. Less: agency rebate (preferably LESS THAN 50%, otherwise it gives them the idea that they are employers)
c. Equals: agency rate (or, better, ‘fee’: always remind those often non-translators that this is INTELLECTUAL WORK, not just ‘amateur bilingual typing’).
—
So AGENCIES MUST CHOOSE:
– either they get the FIXED AGENCY REBATE (e.g. -30% over the end-customer rate)
– or they get the FUZZY MATCH REBATE,
BUT NOT BOTH:
AGENCIES CANNOT GET A FUZZY MATCH REBATE OVER THE AGENCY RATE, SINCE THE AGENCY RATE IS ALREADY A REBATE.
So, I repeat: it’s one or the other – BUT NOT BOTH!!!
Otherwise it is just NOT FEASIBLE. This system has been invented by NON-translators, i.e. commercial people WHO DO NOT BELONG HERE.
2. And if intermediaries were given 10-15% rebates over the end-customer rate, they would know where they belong, i.e. sort of business finders.
Whereas THE PRESENT LSP BUSINESS MODEL is an attempt to become an UNDECLARED EMPLOYER, by forcing freelance translators to use their online CAT tool, under the agency’s orders and supervision, i.e. under their AUTHORITY – sometimes even with real-time monitoring of translation progress! – which corresponds to the very definition of EMPLOYERS – which entails they should be paying for translators’ social security contributions and taxes!
I am very much in favour of a CLASS ACTION against all those large LSPs who FORCE freelancers to use such or such tool, particularly ON THEIR OWN WEBSITES, and, even worse, with REAL-TIME MONITORING of translation progress – like the translated.net platform (which catches customers by heavily advertising on top of Google translation agency search results, with the money they did not pay to translators!) and Xplanation (the largest Belgian agency in 2015 sales figures!) and end-customers, like KBC, Belgium’s 3rd bank, that employs freelance translators half-time (with fixed schedules), on KBC’s premises, using KBC’s CAT tool, and under KBC staff supervision!
THESE ABUSES MUST STOP!
I am sure that with the economic crisis, our governments will be more than happy to get supplementary revenues from all those huge end-clients and intermediaries who act like employers while attempting to avoid paying for social security contributions and taxes!
3. Last but not least, it must be denounced on public platforms that CAT tools CONSIDERABLY SLOW US DOWN AND CONSIDERABLY DECREASE QUALITY.
A 2012 Proz poll among freelance translators showed that ONLY 60% of freelance translators were UNDER THE IMPRESSION (!) that CAT tools increased their (translating or overall?…) speed (compared to what: using MS Word alone, without using the AutoCorrect macros as typing accelerators and glossaries, extremely quick to encode, contrary to the hyper-segmented termbases conceived for CAT tools?):
FAR FROM UNANIMITY THUS!!!
And SDL’s latest gimmicks (dozens of mostly useless and time-consuming application plugins that make their latest CAT tool version most unstable) do not change anything to this.
In fact, SDL’s 2011 termbases are totally INCOMPATIBLE with the 2017 CAT tool version: index cards (‘entries’ as they call them) keep disappearing without warning and for no apparent reason, after having been created or just updated, wasting HOURS OF WORK.
And either SDL crooks pretend they ‘do not understand what’s happening’ or that they ‘don’t know how to solve the problem’.
I.e. SDL failed to mention that the 2011 and 2017 versions were incompatible, but they took our money for the upgrade (some €300).
They have no other outlet for angry customers to vent their frustration… than an online forum… where JUST ANY translator can come and add his or her pinch of (useless) salt and make things worse by STEALING FURTHER FREELANCER’S TIME.
SDL PEOPLE ARE TOTAL CROOKS.
PLEASE LET THE WORLD KNOW….. THANK YOU.
And do not let those non-translating LSPs keep on DISINFORMING OUR DIRECT CUSTOMERS about CAT tools supposedly increasing speed and quality while decreasing prices.
Use the opportunity of your marketing emailings to INFORM YOUR POTENTIAL DIRECT CUSTOMERS.
End-customers HAVE NO IDEA how many words can be translated per hour and thus HOW MUCH they should pay!
If we translators do not inform them, they believe non-translating agencies’ LIES.
I have recently read an LSP organisation’s blog hoping that translators would translate up to… 10,000 source words per hour in the near future:
NO BRAIN CAN PROCESS 10,000 NOTIONS IN JUST 60 MINUTES!!!
It’s UP TO US TO INFORM THEM.
Amen.
Thank you. 🙂
LikeLike
By: Isabelle F. Brucher, Legal, Finance, Marketing & HRM Translator, NL, EN, ES, DE >FR since 2004. on June 24, 2017
at 11:49 am
“Whereas THE PRESENT LSP BUSINESS MODEL is an attempt to become an UNDECLARED EMPLOYER, by forcing freelance translators to use their online CAT tool, under the agency’s orders and supervision, i.e. under their AUTHORITY – sometimes even with real-time monitoring of translation progress! – which corresponds to the very definition of EMPLOYERS – which entails they should be paying for translators’ social security contributions and taxes!”
I don’t agree with everything you say, Isabelle, but I do agree that the new agency model turns translators who were formerly treated as freelancers into de facto undeclared employees. What the new translation agency model has created definitely meets the criteria that the IRS is using to distinguish freelance workers from employess as the requirements on these undeclared de facto employees, who have not benefits and who must pay their own taxes, are often more onerous than those imposed on real and declared employees whose taxes must be regularly paid by their employers.
But I don’t think that a class action is a viable solution. Even if there were one, the lawyers would most likely end up with most of the money.
I do wonder, though – will the tax authorities, such as IRS here in US, start auditing major translation agencies to impose penalties and request past-due taxes from them for failing to pay taxes for undeclared employees?
I know that this is happening in other branches where mostly freelancers are used (IBM was in the news because they were treating de facto employees as freelancers).
I never thought I would say something like this, but here it is: it appears the one hope that translators still have in their fight against the new exploitative agency model is now the Internal Revenue Service.
LikeLiked by 1 person
By: patenttranslator on June 24, 2017
at 12:06 pm