Tuesday, October 19, 2010


“What I get from "traduttore-traditore" is that it is near-impossible to render a text -from one language into another with 100 percent accuracy, because there are always nuances and subtleties in any language that cannot easily be conveyed in another. (If every language had every possible subtlety within it, it would almost be unnecessary to have other languages at all, because they would have the same power of expression.) All but the simplest texts have shades of meaning that are hard to translate with precision.  A sophisticated text can contain ambiguities, intentional and unintentional, and the task of the translator in this case is to convey the ambiguities that are in the text without introducing any that are not.  In reality, the translator's job is to minimize the losses during the passage from one language to another; in other words, the object is to minimize the betrayal!”. 
            - Christine Quinones

Part 2

By Bernie Bierman

To the older practitioners of the industry, the greater of the surprises was not so much the speed with which the change occurred, namely the conversion from what could be called cerebral translation to what could be called robotic translation, but the overwhelming acceptance of the latter by the next generation of practitioners.  While the older group of practitioners most certainly accepted, used and even welcomed the new technologies, there was an element of selectiveness in the process.  One technology that was not accepted – and indeed scorned and resisted – was computerized translation…in any of its varied forms and shapes.

The younger generation did not see the world of translation through the eyes of their seniors or even their mentors.  The younger group had been literally baptized in the rivers, streams, lakes and even oceans of technology, particularly computer technology.  They were fully wired, wireless-ed and gadgetized. The older translation practitioners were fish who had to learn to swim.  To their offspring, swimming was as natural as natural could get. 

But the technological changes in the translation process that began as a trickle in the late 1980’s, morphed into a quick-flowing stream in the middle-to-late 1990’s and became a torrent in the first decade of the new century, weren’t by any stretch of the imagination the sole reason for the massive changes in the industry that were witnessed in the first decade of the 21st century.  The technological changes that affected how we translate and how we approach and view translation were accompanied – or perhaps slightly preceded – by major (if not earth-moving) non-translation-related technological advances that in turn caused or certainly helped bring about major (and again, if not earth-moving) market and economic changes. 

In geologic time, the development of the internet was about a nanosecond.  In terms of the lifetime of the human being, it was about overnight, or for the more conservative, two overnights.  The Internet made next-door neighbors of New York and Paris, Beijing and Berlin, Moscow and Buenos Aires and (not “or”) made communication between Yellowknife, NWT and Novosibirsk as easy as communication between two towns 5 minutes distant from each other.  The oceans separating the continents literally dried up.  And we gave that drying up a name: globalization.  In the first part of the 1990’s, we could still talk about a French translation industry or a German translation industry or a U.S. translation industry or a Japanese translation industry or even a budding Chinese translation industry (as China’s communist leaders were beginning to see, appreciate and even slightly worship good old capitalism).  By the time we were celebrating the millennium (amidst stomach-churning fears that our computers would crash and burn), there was really only one translation industry: the world translation industry.

The last decade of the 20th century also marked the beginning of what would be the very rapid demise (and extinction) of a species I called the “translator-merchant” and the coming of the businessperson (more often than not armed with an MBA diploma in marketing) fascinated with the seeming profit potential of the translation business, a business that appeared to these business types as virgin territory for the implementation of all the theories learned in graduate business school.  One of the more interesting things about the new translation businessperson of the 21st century was that he or she had only the most rudimentary knowledge of language, knew virtually next-to-nothing about translation and the translation process, and appeared in more cases than not, to have little knowledge of and skill in his or her native tongue.  Nonetheless, the translation business impresario of the new century came adorned with the titles of station: “President & CEO”, “Executive Vice-President for Marketing”, “Senior Vice-President of Linguistic Systems Analysis”, ”Translation Resources Coordinating Manager” and all sorts and manner of other titles that could easily bring one to a thunderous commercial orgasm.  But it was the nomenclature applied to this new breed by the late Robert Addis that was really far more descriptively accurate: “Language-blind administrators”.

Unlike their predecessor “translator-merchants”, these language-blind administrators and business impresarios did not hesitate for a moment to avail themselves of either machine translation or its close relative, computer-assisted translation.  Indeed, computerization and robotization of translation were very much in line with the tenets and theories learned in graduate business school respecting productive efficiency and cost-effective efficiency.  Computer-assisted translation in particular was not the future.  It was now.  CAT would eliminate what had been the historical bane of the translation industry: labor-intensiveness.  It would not only do away with labor-intensiveness, but it would definitively allow for greater production, more efficient production and product consistency, for after all, “once you have translated a sentence or passage, you will never have to translate that sentence of passage ever again”, as SDL Trados, a leading CAT manufacturer, promised the translation world.  One of the many proofs of this pudding was the proud (and loud) announcement by Marian Greenfield, a U.S.-based translator and former president of the American Translators Association, that she had translated 33,000 words in 10 hours thanks to CAT (not coincidentally SDL Trados’ CAT). 

Surprisingly (at least to this observer), the European translation world went quickly and overwhelmingly for CAT, far more quickly and overwhelmingly than North America, where at least at the outset there was some discernible hesitation.

By the year 2001, the economic impact of CAT (and to a slightly lesser extent, MT) was beginning to be seen, if not felt.  However, there were two more factors the impact of which would literally turn the translation world on its head.  One was the full-blown globalization of the industry and the other were the economic downturn twins: the more benign one of 2001 and the far more malignant one of 2008, the latter downturn (or if the reader prefers, recession) being one that has not responded too well to even high dosages of economic radiation and chemotherapy.

Globalization and the two economic downturns sent overall translation prices in the U.S. down to late 1950 levels and in western Europe to late 1960 levels.  India and China became major players in this new fully globalized translation world, and competing with Indian and Chinese (and southeast Asian) prices became almost impossible, especially in nations with much higher costs-of-living.  But CAT and its broad acceptance by both translation service providers and their freelance sub-contractors – now more CAT workers or operators than actual translators or linguists – helped not only to push prices downward, but also to maintain them at incredibly low levels. 

But CAT has done something else, something of far great significance than was done by either globalization or the mini-recession of 2001 or the great recession of 2008.  CAT has done away with the concept of translation as a communications service (a communications service requiring some very unique and singular skills) and converted it into a service based solely upon the providing of words, i.e., the sale of words.  The myriad of intrinsic and extrinsic factors that went into determining – many times on an ad hoc basis -  a fee or rate or price for a translation service in pre-CAT days have been dispatched to the proverbial city garbage dump.  CAT programs are able to determine everything “scientifically”, from “new” words to “old” words to “similar” words” to “identical phrases” to “somewhat identical phrases”.  According to conventional wisdom as furnished by the purveyors of CAT tools, the computational power of CAT programs can determine with unquestioned accuracy the percentages of phrases, parts of phrases, sentences, parts of sentences that all fall under certain matching criteria.  Percentages of such matches can be determined with allegedly pinpoint accuracy, with little room for dispute.

The entire process that has been engendered by CAT for evaluating the service is based solely and exclusively upon words.  Just words.  Only words.  Exclusively words.  Nothing else but words.  CAT has converted translation from a communications service into a commodities service, with the commodity being the word, the individual word, the isolated word.  CAT and its champions, proponents, disciples, propagandists and the hordes of happy CAT workers and operators have in this post-translation world brought us “Words for Sale”.

[In Part 3 of this series, the author will detail some of the word-counting acrobatics performed by CAT tools that have resulted in translation having become a words-for-sale commodity, and at the same time converted translators into ordinary robotics workers…at corresponding compensation.]

Saturday, October 2, 2010


“I am sure you are not following the trend.
If you are frozen in the 70s and want to keep working
with your typing machine there is nothing I can do”.
- Renato Cedin
  Link Translations (Brazil)

“Welcome to the 21st century: Computers are the future of translation
If you want to be as efficient a translator as possible, Microsoft Word and Excel
should not be your primary translation environments”.
- Jon Ritzdorf, Solutions Architect
  Introduction to Computer-Assisted Translation
  New York University - SCPS

CAT for sale
Words perfect and unspoiled
Words never even slightly soiled
CAT for sale

Let the translators pipe of words
In their now extinct way
CAT knows every type of word
Better far than they ...
- Inspired by Cole Porter’s “Love For Sale”

(A Multi-Part Series)
Part 1

By Bernie Bierman

As far back as my memory can take me (and that is indeed many years), translators have been debating and arguing the question of the value of translation, and more particularly the best or most effective way to calculate that value.

My own memories and written records of this debate and argument reveal that various approaches were tried over the years in an attempt to calculate that value: per line, per word, per typestroke, per hour, by exotic formulas that combined the per line, per word and per hour criteria, and by all sorts and manner of esoteric calculations. 

For reasons that are still not clear, the per word method outlasted and outlived all the others.  Although many translation practitioners decried the per word calculation as inadequate, the consensus was that it was the most adequate of all the inadequates.  However, it did have the financially-redeeming virtue of consistency, that is, if one could agree on what precisely was a “word” for purposes of calculating a price.  Was a number in digital form a “word”?  And if the number in digital form was viewed or counted as a “word”, was 1000 really two “words” rather than one “word”?  Was 1,789,348 one “word” or nine “words”?   Was a name a “word”?  If the Italians called their city “Firenze”, was the English-language version of “Florence” a translated “word”?   And in the pre-computer and pre-CAT age, if names and numbers were viewed as countable or calculable “words”, the question or problem often arose as to how one precisely went about the actual counting of words?   Did the document really contain 8600 words, or was the real count 8492 or 8710 or 8628? 

And if the question of calculation and/or the method of calculation weren’t sufficient enough to stir the arguments, there was the matter of which words should form the basis of calculation: the source language words or the target language words.  Which was economically more equitable to the translator?  Which was economically more equitable to the client or user of the translation?  Was the source language calculation an accurate reflection of the overall translation process?  Or was the target language calculation a much more accurate reflection of that overall translation process?  And on and on the debate went…and raged.

* * * * *
Here in the United States, a man named Lewis Bertrand came upon the commercial-industrial translation scene in the 1920’s.  Bertrand was not merely a translator and writer (which has the same redundancy as “composer and musician”), but also a highly astute and (soon-to-be) successful businessman in the translation industry.  By the 1930’s Bertrand’s influence on the tiny U.S. commercial-industrial translation community was quite discernible and that influence grew in the following two decades.

Indeed, Bertrand’s incessant preaching of his gospel of translation economics brought many converts and spawned many faithful.  Henry Fischbach, the co-founder of the American Translators Association was a disciple of Bertrand’s translation economics, along with so many of the men and women who helped found and build what is arguably one of the largest translator organizations in the world.  (Sadly to a few, the leaders and members of the modern ATA haven’t the foggiest idea of who this Lewis Bertrand was, but even more sadly and more regrettably, the concepts developed by him respecting the essence and value of translation as a commercial endeavor and later promoted and fostered by such founding personages as Henry Fischbach, the Mins brothers, Robert Addis and so many others whose names are now long-forgotten, have been discarded like a piece of rusting machinery as a new generation of computer worshippers and technological “geek-speakers” have come along to proselytize and spread their own gospel.)

In economic and commercial terms, translation to Lewis Bertrand was not about words.  The idea that translation was about the sale of words was so anathema to him that he could literally break into a wild rage over the slightest suggestion in that direction.  To Bertrand, translation was about communication; translation, particularly in the commercial and industrial domains, was a communications service; translation was not about the supply of words; translation was about the supply of ideas, concepts, knowledge, a supply delivered by a person or persons having certain well-defined and unique skills…and talents.

I was only a teenager when I first heard one of the numerous Lewis Bertrand stories about translation and translation economics.  The one particular story that remained with me for a lifetime – as if it had been branded into my brain – went like this:

He was asked by one of his firm’s more difficult and chronically-complaining clients why it was being charged for “translating” names and numbers and what was the rationale for this “seemingly unfair charge”.  He was asked by this difficult and chronically-complaining client whether his firm could eliminate all the names and numbers from a translated document, thereby lowering the invoice by what could be a considerable sum.

Bertrand listened attentively to his client’s complaints and attempted to explain the nature and essence of a translation service.  But the client was adamant.  It did not need names and numbers “translated” and certainly did not want to pay for any such needless “translation”. Bertrand politely acceded to the client’s wishes.  The next translation job to be done for this client would be absent all non-translation elements, i.e., names, numbers and other “noise of like purport and tenor”, and the invoice would duly reflect a sharp reduction in the client’s translation cost.  The result was something like this:

“Dear Mr. [see original] :

“We acknowledge receipt of your letter dated [see original] and with regard to the delivery of [see original] tons of [see original] rolled steel scheduled for shipment on [see original] via the SS [see original], we would like to advise you as follows:

“Because of major dredging problems in the port of [see original], Captain [see original] of the Port Authority of [see original] issued Directive No. [see original] which prohibits vessels over [see original] tons displacement from entering port channels Nos. [see original] and [see original] until [see original] at the very earliest.  Although we could arrange for transshipment in [see original] between the dates of [see original] and [see original] to a vessel of lesser displacement, the costs would increase, and the letter of credit No. [see original] issued by the [see original] Bank in the amount of US$[see original] would be insufficient.

“Should you have any further concerns, please feel free to contact the undersigned by telephone at [see original] or telex at [see original], or his deputy manager, Mr. [see original], whose telephone number is [see original] and whose telex number is [see original].

“Yours, etc.

“[see original]

By: [see original]”

The upshot of this was a very quick telephone call from the chronically-complaining client: “Mr. Bertrand, we got your point.  Please provide us with a complete translation and charge us accordingly”.

Lewis Bertrand was never happy with or enamored of the per word method that was the dominant method in calculating the value of a translation service.  He saw it as a reference tool at best, a reference tool that was far too often not reflective of the true value of the service and all of the elements that went into providing the service. At worst, he viewed the per word invoicing method as a manifestation that the translator was really selling just words.    He and his contemporaries and competitors often tried combinations of various calculation methods, such as per word and per hour or various complex formulas that combined numerous calculation methods.  Bertrand believed fervently and passionately that translators possessed certain highly-defined skills and knowledge that could not be adequately compensated by the mere counting of words and attaching a figure to that count. 

“We are in the communications business”, he would often bellow.  “We are not a bunch of (expletive) clerks”, he would yell.  “We are not  (expletive) typists or transcribers”, he would scream.  [In the mid-1940’s, he charged a client US$100 to translate one word – it came out as two words in the target language.  “If an advertising agency can charge a client thousands of dollars to come up with a 5-word slogan or phrase”, he said, “then surely all of the thought, time and talent that went into finding the right expression in another language for that one English word makes the $100 fee not just a bargain, but a steal”.]

In his day, lots of people in the translation industry (both in the U.S. and Europe) listened to this fiery preacher of translation economics, and many followed his preachings.  Although Bertrand was certainly around in the late 1950’s and early 1960’s, when the first attempts at computerized translation were undertaken, he and so many of his contemporaries dismissed this endeavor as a bad joke and as a need by certain people to robotize everything in sight (and even out of sight).

Clearly, if some “geek-speaker” appeared before a vast assembly of translators in the days of Lewis Bertrand and his contemporaries – or even in the days of their successors – and told the audience that “once a word or sentence or phrase or paragraph had been translated, that very same word or sentence or phrase or paragraph (or a close facsimile thereof) would never ever have to be translated again”, the laughter and howls in the assembly hall would have been of a decibel level that would have surely compromised the structural integrity of that assembly hall.

If Bertrand and virtually all of his contemporaries were blind to the coming of robotized translation, which was not merely on the horizon, but rather just around the corner, there was not so much as a nanosecond thought given to the possible economic impact upon the translation industry that this robotization would wrought.   Of course, there was only handful of people anywhere who saw a world connected through a place that didn’t as yet even have a name.  And never mind the word “globalization”, the concept alone was if not science-fiction, then certainly economics-fiction. 

[To be continued in Part 2]