Monday, August 13, 2012

What is Professional Translation?: The Quality of Smartling's Spanish Website

Some men are born mediocre, some men achieve mediocrity, and some men have mediocrity thrust upon them. With Major Major it had been all three. Even among men lacking all distinction he inevitably stood out as a man lacking more distinction than all the rest, and people who met him were always impressed by how unimpressive he was.
― Joseph Heller, Catch-22

And, sure, he is an honourable man.

Julius Caesar, Act III, Scene 2

If you recall, a couple of months back I had a curious experience when Pinterest called for its users to crowdsource the Spanish version of the site. The thing was that the blog post the company used to energize its crowd was not so much in Spanish but rather in what becomes of Spanish after a bloodthirsty psychopath chops it up into itty bitty pieces, stuffs the remains into the trunk of his car and drives away. The Pinterest employee behind this monstrosity insisted it was perpetrated by a professional translator. I countered by saying: “No uh.” And she finally relented and admitted that her mother had done the translations (although she was careful to delete the smoking gun tweet in which she attributed the work to her mom). Anyway, my narration of this ridiculous affair was crowned by a recollection of a similar incident when a crowdsourced “t9n” company called Smartling proudly announced the launch of its Spanish-language website. The funny thing is that what Smartling calls “Spanish” is not so much strongly influenced by the tongue that emerged when the Angles met the Saxons met the Normans. No, it actually is the tongue that emerged when the Angles met the Saxons met the Normans. I tweeted the fact that Smartling’s “Spanish” website was actually in “English” (which is, like, a whole other language). This prompted frantic tweets from an employee asking what the problem was. After I informed her, she equally frantically rushed to put up some sort of Spanish version online.

As I described elsewhere, this two-sentence comparison between Pinterest and Smartling prompted a backlash from the very irate chief executive officer of the latter company, Jack Welde. In his rebuttal of my criticism of crowdsourcing, he stated that “there is plenty of work for professional translators, especially the good ones. And Smartling is delighted to work with some of the best translators in the business; we respect their craft and the high quality work they do.” Earlier, he had noted that “many of our customers use professional translators to perform translation -- translators like yourself (although you seem pretty angry, and not much fun to work with...).” Anyway, trollish comments aside, I did promise that I would publish a slightly more detailed appraisal of 1) Smartling’s own Spanish language website, (which I suppose would have been assigned to these “professional translators” Welde claims to work with) and 2) a sampling of the websites of Smartling’s own clients.

Let us begin by recalling the main highlights of the Welde Translation Philosophy. He is quick to stress that for technical materials, crowdsourcing is not appropriate:

Would we recommend crowdsourcing the translation of legal content, highly technical materials, or financial content? Nope, we would recommend professional translation from translators skilled in that vertical -- perhaps someone like you... But for companies with a passionate community of users who know the product or service intimately, crowdsourcing translation using high-quality tools to manage the translation process among a large group of participants can be a terrific way to increase community engagement -- and typically with much faster turnaround.

In his view, crowdsourcing is ideal for social media purposes. The other main pillar of his pitch (and also heard often) is that crowdsourcing is not done to save money, but rather to enhance users’ engagement with the platform:

It's generally not about "the money". I'm pretty sure Pinterest can afford to pay for professional translation, but I suspect they are looking to incorporate their existing passionate community into the translation process as a means of increasing engagement -- while moving at the speed of Web 2.0 businesses.

The general message is that Smartling’s platform is agnostic and neutral. You can localize your website using an agency, in-house translators or your website’s users. I assume that Smartling’s own website was translated using these much-vaunted professionals. Listen to the CEO extolling the output of the professionals he employs: “Smartling is delighted to work with some of the best translators in the business; we respect their craft and the high quality work they do.” Now look back at the quotes from the Smartling boss and see how many times the highlighted phrase "professional translators" pops up. It is obviously an important part of his pitch. It is reasonable to expect that proof of the high quality provided by these translators can be found in the face that Smartling presents to its Spanish visitors, I imagine. So let’s return to the scene of the original crime. Let’s click on the language tab of Smartling’s home page and go through the looking-glass.

In my view, translation is something that can be done by any bilingual, with differing levels of success. Professional translation, in contrast, is the product of thought applied to the everyday task of translation. Viewed under that light, it is readily evident that Smartling does not employ professionals even for its own website, since very little real thought has gone into the work. It is not so much that Smartling’s bilinguals are incompetent, but rather that they do not have any experience in the difficult task of laboring over a message in one language and then coming up with an equivalent in another one. And that is why the translations Smartling facilitates for itself and its clients sound a little like the end-of-year project completed by heavily stoned middle schoolers for their Spanish 101 required credit.

Look, for instance, at the website’s menu. “Traducción de la comunidad” as an option for “Community translation” is wrong. To give you an idea of how wrong it is, when you back translate it, you get “Translation of the community.” “Traducción comunitaria” would be a better option. “Kit de medios” as an option for “Media kit” is just embarrassing. 

A site menu is an object to which you devote a lot of thinking, because it determines how visitors surf your web page. It may be just 20-25 words, which usually can be translated in a few minutes. But you should devote several hours to choosing the words carefully in order to keep those fickle Internet visitors from being instantly turned off by a stilted and clumsy Spanglish roadmap.

“Factoid” was localized as “factoide,” which is a hallmark of the professionals who graduate from the Taco Bell School for Spanish Translation. Their methodology consists in basically adding an “e” or an “o” to any English word to make it sound like Spanish. The content of the “factoides” themselves are somewhat difficult to figure out. Check out number 1:

¡Con una población de unos 32 millones en 2010, los mexicano-estadounidenses comprenden el 63% de todos los hispanos de EE.UU. y el 10% de toda la población de EE.UU.!

First of all, why the exclamation marks? The idea that a dry statistical fact is worthy of opening with an exclamation mark in Spanish is dumbfounding. Answer: the exclamation marks are there because the original English has one, which is precisely how non-professional translators tend to work. 

Everything in these sentences is clunky, from word choice to the grammatical sequence. The structure of the sentence transcribed above is a carbon copy of the original ("At nearly 32 million in 2010, Mexican-Americans comprise 63% of the U.S. Hispanic population and 10% of the total U.S. population!"). But it is the use of “comprender” for “comprise” that just kills any hope of reading comprehension. There is a bouquet of other word choices that would make a lot more sense and would help the reader more (incidentally, this tends to heighten the suspicion that this text is the product of a cursory post-editing by an inexperienced linguist, but Jack claims emphatically that he doesn’t do post-editing, “and Brutus is an honorable man”). A sentence such as this is the product of either a machine translator or a very unskilled human one, which for all intents and purposes come to be pretty much the same thing.

The same amateurish handiwork is evident in Mr.Welde’s profile page. The literal translation "hombre del renacimiento" as an option for "Renaissance man" is meaningless in Spanish. A professional translator would tell you that. Raw machine translation won't. Neither will a crowd of hamsters. They will also fail to tell you that acronyms as frequent as CEO and MBA have very nifty equivalents in Spanish. 

In the following sentence, the somewhat chaotic profusion of capital letters is once again the product of acritical copying from the English original:

Es licenciado en Ingeniería Informática por la Universidad de Pensilvania, donde también estudió Lingüística e hizo prácticas con el Profesor William Labov, y tiene un MBA de la Universidad de Cameron (Alemania).

And now observe this complete and utter failure to even approximate the English original (He lives outside of NYC with his wife and children and can usually be found writing product specs at midnight, discovering new music or flying light aircraft):

Vive en las afueras de Nueva York con su esposa e hijos, y es fácil verlo escribiendo especificaciones de productos a medianoche, descubriendo nueva música o pilotando aviones ligeros.

Es fácil verlo escribiendo...” That, my friends, is the sound of the post-editor throwing his arms up in despair and screaming: “Screw this! I’m only getting five dollars an hour! Let the proofreader take care of this!” Either way, Welde has some gall to tout his collaboration with professional translators when he, defying belief, doesn’t even use them when his own image is at stake. Here is the back translation:

He lives outside of NYC with his wife and children and it is easy to see him writing product specs at midnight, discovering new music or flying light aircraft.

Why is it so easy to see Jack writing product specs at midnight? Hasn't he heard of walls? Does he do a Big-Brother type webcast of his home life? 

And so on and so on.

A reader called Juliana reported in a comment that the quality of the Portuguese version of Smartling’s site is equally poor:

By the way, I'm from Brazil and decided to check out Smartling's "how it works" section in Portuguese. Of course people will understand what's being said there, but the writing is awkward, clearly unprofessional, garbled even. I don't understand how people can extol the virtues of Web 2.0 and at the same time not give a rat's ass about the quality of their content, since it's all about ENGAGING PEOPLE THROUGH WORDS ON THE SCREEN.

Couldn’t have said it better myself. I hope I have provided enough evidence to prove that if Smartling does indeed use professional translators, it does not use very good ones.

Now, Mr. Welde is free to promote his business as he sees fit. However, his repetitive claim that Smartling employs professional translators should not go unchallenged, because a cursory inspection of his and his clients’ websites clearly demonstrates that he doesn’t. My fear is that Mr. Welde probably does not have any acquaintance with the non-English-speaking world aside from that time in the mid-nineties when he spent a summer bombing Serbia from his laptop. 

His profile claims that he holds “a [sic] M.B.A. from Cameron University (Germany)”. Curiously, the Internet reveals that there is no Cameron University in Germany. There is a Cameron University in Oklahoma, though.

Jon Voight as Milo Minderbinder in the movie version.
Oklahoma. Germany. Different places, in my view. “Same difference,” in Welde’s world view. I shudder to think that this same dude was picking targets during a NATO bombing campaign. If he employed the same geographical acumen in that task that he uses in describing his alma mater, we may have a post-modern version of Catch-22 on our hands.

And, to tell the truth, Welde does remind me a lot of the Lieutenant Milo Minderbinder immortalized by Joseph Heller in his classic satire about World War II. Minderbinder is a red-blooded, blond and blue-eyed officer who runs an illegal bartering operation using matériel he stole from the Air Force. He justifies all his actions by blithely stating that “what's good for M&M Enterprises will be good for the country.” The M in M&M stands for Milo, of course (he added the “&M” so people wouldn’t think it was a one-man operation). In one climactic scene, Yossarian’s plane is going down and he opens his parachute to discover an I.O.U. from Minderbinder, who “borrowed” the parachutes’ silk to make stockings for prostitutes.

Why is Cameron University suddenly transported from the arid badlands of Oklahoma to the lush, green fields of Germany? Is it perhaps because Welde earned an M.B.A. online from Cameron University while living in Germany? That would be my guess. Is this, then, perhaps the case of a slightly unworldly American businessman trying to puff up the international aspects of his CV because he runs a translation company but doesn’t know any other languages? Possibly. Is it insane to point out that this little obfuscation might be somehow related to the low quality of the translations on his own site? Who knows? 

The world is a mysterious place (albeit endlessly fascinating in its sheer absurdity).

(In a future post, I will publish a review of the localized websites of Smartling’s clients to determine the degree of success with which these companies, in Welde’s breathless prose, have incorporated “their existing passionate community into the translation process as a means of increasing engagement -- while moving at the speed of Web 2.0 businesses.”)

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. To contact him, visit his website and write to the address listed there. You can also join his LinkedIn network by visiting the profile or follow him on Twitter.

Thursday, August 9, 2012

‘Reach Down, God, Give Me a High Five’: Low Quality Translation and Low Quality SEO

This guy thought up Death Bed: The Bed That Eats People and f****ng finished it! That means one of two things happened. He either he never had a moment’s doubt. He hit that typewriter every day. "And then the pillow starts to smother… Ohhh! This is awesome! Reach down, God! Give me a high five! Boom!” Here’s what’s worse. What if he had moments of doubt, AND THEN F***ING WORKED THROUGH THEM? That’s so much worse for me. What if he was going: “And then the pillow starts… What the f**k am I writing? I’m putting my name on this piece of… No! I will finish this!” He looked at his poster of the little kitten hanging from the tree saying “Just hang in there, baby.” And he said: “Yes, I will hang in there, kitten.”
—Patton Oswalt, Werewolves and Lollipops

You know one of my bugbears (or perhaps hobby horses) is the Content Tsunami. It is the main pillar of the flimsy business case for Low Quality Translation. It goes a little something like this: “Since the amount of content is exploding, we need low quality translation to translate this flood of (low quality) information.” I want to use this opportunity to highlight one tiny little molecule in the endlessly expanding ocean of the Information Big Bang.

The piece is published in the blog of a Very, Very Large Translation Agency that does a lot of Spanish post-editing at $0.02 per word and constantly badgers qualified professionals to join its ranks of underpaid drones. In the immortal words of Forrest Gump, “stupid is as stupid does.” The blog post I am discussing here is, perhaps uncharacteristically, not the product of a computerized copywriting program. I can safely say it was actually produced by a human being. But, as Low Quality theorists are fond of reminding us, human authorship is no guarantee of quality:

Financial documents can be produced in a variety of file formats. Keeping this in mind, Trusted Translations is prepared to accept all types of files, and can deliver them as ready-to-publish files if so required by the client.

Thank GOD for Trusted Translations! Where would we be without an unscrupulous, faceless corporation and its semi-anonymous ten-dollar-an-hour blogger reminding us that financial documents come in a variety of file formats? Thank GOD for the Internet! To think that as recently as 1993 you couldn’t drive your PC on the information superhighway and come across this banal piece of drivel.

But, as Jon Stewart says, “Wait, there’s more”:

Finance departments, along with financial institutions themselves, are a key area in managing any type of business. Producing documents that hold very important and specialized information, these departments often require accurate translations of these documents in order to communicate financial information to a business’s own offices in another country, or to other companies. Trusted Translations has experience quickly and accurately translating a range of financial documents and has access to resources such as proprietary financial dictionaries, translations memories and expert industry-specific translators.

"Trusted Translations has experience quickly and accurately translating a range of financial documents..." Can you just imagine the anonymous blogger writing this sentence and crying out to God for a high five? Let’s parse this. Proprietary financial dictionaries. Yeah. If you place the search phrase “financial translation” in Google, your first result is a bilingual glossary that purports to be specialized in finance, courtesy of… you guessed it! Trusted Translations, the finest purveyor of Low Quality Translation. The glossary bears the distinguished title of “English Spanish Dictionary of Financial Terms.” And, obviously, it was crafted by a bevy of “expert industry-specific translators” (?), who, I am guessing, are the ultimate arbiters of the text after it has been processed by Trusted’s machine translators, Roombas, C3-POs, Wall*Es, and sundry translation memories. What do these “expert industry-specific translators” consider worthy of including in a financial glossary? Let’s see. “Go-go fund.” Yes, that comes up very often in financial documents… written in 1965. So, if you are ever swallowed up by a worm hole and deposited in the year when I Dream of Jeannie was number one in the Nielsen ratings and Vietnam was a distant place where a handful of Marines were spending the nastiest summer vacation ever, well, golly, Sarge, Trusted Translations just saved you a lot of time!

TT’s contribution to the Content Tsunami is, of course, nothing more than cheap SEO-gaming without bothering to actually contribute anything of any value to the Internet. My thesis is that this opportunistic online marketing ethos is indicative of its overall business philosophy (cheap, cheap, cheap...). Allow me to provide a sampling of the blog post’s internal hyperlinks. The phrase “financial translation” leads the accidental cyber-tourist to a cluster of articles (of similar quality) on issues as diverse as “financial translation teams”, “financial translation languages”, “financial document translations” (and, let’s face it, who hasn’t googled those Boolean phrases in the wee morning hours of some desperate, lonely Saturday night?) 

And so on and so on. The thing that gets me is that TT positively RULES the search rankings. Not only does it broadcast its low quality content in every single localization conference, it also dominates the online search world with the same iron fist with which Ivan the Terrible ruled early modern Russia.

“Stupid is as stupid does.” If your translation provider uses Low Quality search engine optimization, what are the odds that it doesn’t use Low Quality Translation? And passes it off as the work of “expert industry-specific translators”? Hmmmm…

Trusted Translation’s SEO strategy is just the same adolescent hacker ethos that underlies Low Quality Translation, made even more grotesque by the fact that it is espoused not by teenage computer nerds who don’t know any better but middle-aged gurus who should. Which leads to an interesting observation. When they talk to translators, lower quality translation providers preach the necessity of low quality translation (and, implicitly, correspondingly low rates). But when they talk to clients, these same companies masquerade as high quality translation providers. They mumble in their clumsy corporate prose about “expert industry-specific translators.” They bloviate about their knowledgeable post-editors. Meanwhile, these selfsame post-editors are in a nearby supermarket check-out line trying to pay for baby formula with food stamps, praying to Yahweh and Harry Reid that the Republican Congress will extend welfare benefits for another six months. 

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. 
To contact him, visit his website and write to the address listed there. You can also join his LinkedIn network by visiting his profile or you can follow him on Twitter.

Saturday, July 21, 2012

How to Keep the Debate on Translation Pricing From Going Full Retard

Kirk Lazarus: Everybody knows you never go full retard.
Tugg Speedman: What do you mean?
Kirk Lazarus: Check it out. Dustin Hoffman, 'Rain Man,' look retarded, act retarded, not retarded. Counted toothpicks, cheated cards. Autistic, sho'. Not retarded. You know Tom Hanks, 'Forrest Gump.' Slow, yes. Retarded, maybe. Braces on his legs. But he charmed the pants off Nixon and won a ping-pong competition. That ain't retarded. Peter Sellers, "Being There." Infantile, yes. Retarded, no. You went full retard, man. Never go full retard. You don't buy that? Ask Sean Penn, 2001, "I Am Sam." Remember? Went full retard, went home empty handed...
Tropic Thunder (2008)

It seems as though the Common Sense Advisory has launched a rate survey in order to imbue the idiot-boy shouting match about falling translation prices with some empirical content. The survey has come in for a lot of criticism just because of its formulation. I am not a statistician, but I do know that the creation of an unbiased set of questions is not a simple thing. Kevin Lossner, one of the voices who doubt that prices are declining, regularly summarizes translation surveys by sector organizations that do not point towards the Götterdämmerung celebrated by many self-appointed gurus. His conclusion is as follows:

Rate surveys without a particular ideological or commercial agenda from the ITI and IoL in the United Kingdom and the German BDÜ have larger samples of service providers in many languages, and they mostly tell a story of stable or slightly increasing rates for language service providers.

This to me sounds far more likely than any other sort of scenario (that is, if you are seriously interested in common sense). Of course, there is also the probable fact that translation pricing (mostly for the B2B sector) took a cliff dive in 2008 and 2009, but that is surely due more to the Great Recession than to any factor endogenous to the translation industry (recessions tend to put pressure on prices of all sorts of things because demand craters).

Now, in a vain attempt to keep the debate from going full retard, I have a couple of observations. First of all, even before the current crisis, we lived in a period that some analysts described as the Great Moderation. What does this mean? For reasons that are not altogether clear, the past 30-odd years have been marked by a dramatic cooling off of inflation worldwide. Whereas the seventies were dominated by a precipitous plunge of the US dollar against the ounce of gold, several oil price shocks and runaway inflation, the next three decades were marked by ever lower inflation. Former Fed chief Alan Greenspan recounts in his memoirs how he used to get up every morning and look at himself in amazement in the mirror: productivity and economic growth were skyrocketing, but there wasn’t even a hint of inflation. Many hypotheses were proposed by our dismal scientists (i.e., economists) to explain this remarkable state of affairs. The two most popular explanations were: China (cheap labor) and technology (greater productivity with equal inputs). The latter explanation blossomed further into the idea of the New Economy in the 90s. Basically, it stated that the Internet and better communications technology were allowing the world economy to achieve runaway growth without inflationary pressures. Oh, one corollary of this brave, new theory was that recessions were a thing of the past. Was the “New Economy” theory correct? Well, three little incidents happened along the road to recession-free Tech Utopia: the 2000 bursting of the 90s Internet bubble; the 2001 recession; and the worldwide doo-doo storm known as the 2008 financial crisis (and the ensuing Great Recession). That is why I find it hilarious to see the translation gurus peddling the same outdated ideas about a tech-driven deflationary spiral a full decade after the theory was totally discredited by a little thing called reality.

Even now, four years after the crisis began, the thing that keeps central bankers awake at night is deflation. The financial industry has an in-built bias against low interest rates: holders of financial assets hate them. But central bankers in both the First World and the Third World are still fretting about the opposite phenomenon: their nightmare is about the entire world going down a deflationary drain. If you think lower and lower prices are neat-o, e-mail your closest Japanese acquaintance and ask him how deflation has been working out for the Land of the Rising Sun. If you suggested to any of those central bankers that technology has anything to do with our current problems, he would laugh in your face, because obviously the factor putting pressure on prices currently is fear. Good old fashioned fear. Right now, if you buy a German short-term bund, it gives you a negative interest. Yes, you get less money back when you invest in some German and British tenors. Investors are basically paying safe-haven countries to hold their money. There is so much fear around that the smart money prefers to forgo interests and simply pay a small fee to have a creditor country safeguard their money.

Okay, now for the second observation. In a context of low inflation or even deflation, prices that remain stable or increase slightly are actually posting huge increases in value as everything else drops in monetary value. So, any discussion of higher or lower translation prices is pointless outside the macro context of where prices are going worldwide. For now, world prices are stagnant and threatening to shrink. In that context, stable to slightly higher prices mean that you are winning the rat race. If that is your case, pat yourself on the back. If not, try to find some way to get on the winning wagon.

Remember: fear and unfounded paranoia are only certain recipes for going full retard. And, as Sean Penn discovered, that never pays off. As Robert Downey Jr. reminds us: "Never go full retard."

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 SpainTo 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.

Thursday, June 28, 2012

Did You Hear That? Translation Prices Are Falling!

Is that cannon fire or is that my heart pounding?
--Casablanca (1942)

Is that the thunder of distant guns or is it the sound of translation prices crashing to the ground?

Wait, no. It’s only the gruesome sound of McLSPs biting, kicking and eating each other alive as part of a massive cannibal apocalypse.

And the bloodcurdling squeals of a million hamsters…

What a dreadful sound!

This is how groupthink works. Nonsense Common Sense Advisory CEO Don De Palma decides that translation prices are falling because of the impact of automation and all those nifty forty-year-old technologies. In an e-mail to one agency owner, he writes:
Even though the industry has reported strong growth overall each year, our previous pricing survey showed that rates went down for nearly every language between 2008 and 2010. Have rates decreased even further from 2010 to 2012? Or, are they starting to stabilize for some languages?
His clients in the Cheap Translation sector shout: “That’s true! I get dozens of unsolicited CVs from completely unqualified people every day!” David Grunwald, for instance, goes on to conclude the following:

Workflows involving MT are being used more-and-more by LSPs and translation buyers. This is cutting many translators out of the loop, causing a glut in the supply of human translation resources. At GTS, we receive  hundreds of applications a month from under-employed translators.

Now, please note two things:

1.- De Palma didn’t actually say prices are dropping. He noted that prices dropped in 2008-2010, but that corresponds to the severest portion of the deepest worldwide economic downturn since the Great Depression. He doesn’t actually say that prices have dropped since then, either. He has equivocated on the issue of prices over the past two years, but in most cases he tends not to cite any concrete empirical evidence that tends to confirm his sweeping observations about prices (or about much of anything, for that matter).

2.- Grunwald adds some empirical evidence: the abundance of CVs from under-employed “translators.” (I can safely say that Grunwald’s definition of under-employed translators is different from mine. I may concede the “under-employed” part, although not perhaps the “translator” part.) For my part, even though my website explicitly states that I am not an agency and I carefully cultivate a gruff, grumpy public persona, I get dozens of CVs from clueless translators-cum-spammers regularly (by the way, if you’re reading this, I regularly mark your messages as spam, which further decreases the likelihood that your e-mails will reach any real clients). I do not think that says much about the “market” but rather about current Internet culture, which tends towards cheap communication, a model that Grunwald is probably better acquainted with than me.

Furthermore, despite the fact that Spain is undergoing a deep, secular recession, I just had the busiest month for a long time and one of the best months ever from the point of view of revenue. Does that mean I think that human translation is booming? No. That is only an isolated data point in a sea of data points. Worse, it is just unstructured anecdotal and highly biased evidence, which is the basis for 90% of De Palma and Grunwald’s outlook.

People need to learn to think critically. Even numbers into which society invests a lot of effort are just vague approximations. Few people know that a figure such as the US jobs number has a margin of error of plus or minus 100,000 jobs or that it is revised continuously for several months after it is released. The quarterly and yearly GDP numbers, likewise, are constantly revised for many months and even years after they are announced. And those are two key figures in which millions of dollars are invested and which depend upon the work of thousands of survey takers, economists, and statisticians. One of the reasons why people should study economics is to be less impressed by the “reality” of big-sounding numbers. Most of the numbers bandied about an industry as tiny as translation are little more than fluff on some geezer’s spreadsheet. And often even less than that.

Wednesday, June 27, 2012

The Localization Industry Doesn’t Get the Local Web

I heard ten thousand whispering and nobody listening.

We are so constantly bombarded by the ideology that the “world has been flattened” by the Internet that any contrary evidence gets short shrift. The Economist some months ago carried a short review of a study from the OECD and Boston Consulting Group on the global Internet that will surely not receive any play among the Web 2.0 hypers and l10n ideologues. The main conclusion is that “the global ‘network of networks’ is shaped by local forces.” Instead of homogenization, you have a plurality of cultural approaches and uses of the Internet:

the Internet will continue to become more and more local: cultures are different, so the more people go online, the more the Internet will resemble them. “There will be hundreds of internet flavours,” he says.

I can see that pretty readily in visiting, for example, the Spanish blogosphere. Very few people are using SEO to bring traffic in from search engines. No one that I know of is planning to monetize their blog. Very few companies or individuals use them as a Trojan horse to sell other products. The translation blogosphere is very active and lively. Here in Spain, every single translator has a blog. Even T&I students come out with their blogs and they get dozens of comments. I imagine that Spanish profs at T&I departments are telling their students that it is a good way to make yourself get known prior to graduation, but I don’t think professional advancement is the main motivator for most. However, the function of these blogs is completely different. These blogs are, in my opinion, an adjunct to social media. It could sort of be considered networking, but a type of networking in which the personal and professional are not as distinct as in the U.S. The scene is large and chaotic, but also very dynamic.

To take another example, Kaiser Kuo, a spokesman for Baidu who lives in Beijing (and is featured in this week’s episode of This American Life) warns that the hysterical idea of a “Westernization” of Chinese media is erroneous. In his view, Chinese Internet culture takes a lot of Western content and turns it into something completely Chinese (and, by the same token, utterly incomprehensible to foreigners):

A lot of the memes that have become popular in China are sort of indecipherable to Western audiences. And, of course, that is largely because they are irreducibly Chinese. So I think the idea that Chinese culture is becoming westernized is a little misguided. I don’t think there’s a strict dichotomy between Western and Chinese culture.

The idea Kuo is referring to is of an autochthonous Internet culture that is untranslatable. Or—to be more precise (since everything is translatable)—the idea is that Chinese Internet culture does not need translation because it was never meant to be translated in the first place.
The idea of different flavors of Internet culture shouldn’t be such a surprise to an industry that has been banging on (acritically) about localization for well-nigh over a decade. But now, faced with the challenge of the Internet, a lot of the l10n preaching turns out to be a little hollow. The problem is that the Lower Quality Translation Movement runs against the grain of local Internet cultures. This is because, at heart, the type of localization championed by large agencies and its smaller tech competitors is a one-size-fits all model. The ideal is to take any text tailored for an American audience and immediately multiply it into ten thousand languages, like a Gremlin sprinkled with tap water. The concept of one Internet that is localized using cheap translation into every single language on Earth is nothing more than the old model of traditional one-way, Anglo-centric, US-dominated media that produces a standard product which was then “localized” and distributed worldwide.

But perhaps a huge financial crisis and the “Rise of the Rest” open the door for another model. One in which local knowledge is prized above cheap instant translation. In this world, professional translation of commercial texts could be a competitive advantage for smaller players wishing to distinguish themselves from the large multinationals who, on the advice of tech-savvy “localization” providers, just pump out a third-quality translation. The sum total of the expertise provided by l10n players and consultants is little more than “let the Chinglish chips fall where they may” or “let the crowd put a smattering of lipstick on the post-edited pig.” In that world, one company communicates with its local audience in a way the audience recognizes as comprehensible. Simultaneously, ten thousand “localized” websites using MT and crowdsourcing whisper in Chinglish, but no one listens.

I can just hear the ideologue in the background going: “Well, gee, the lack of homogeneity in the ‘global’ Web is actually an opportunity for more homogenization through (subpar) technology and (commoditized) translation.” Please go back to the beginning of this and read it again. (Jesus!) The localization industry simply does not get the local Web.

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. To contact him, visit his website and write to the address listed there. You can also join his LinkedIn network by visiting the profile or follow him on Twitter.