Monday, November 28, 2011

Common Sense Advisory Pummels Defunct Equine Mammal (More on the Content Tsunami… Sigh!)


The Common PlaceSense Advisory—which functions as a sort of propaganda arm for the Lower Quality Translation movement—has jumped on the bandwagon of the Content Tsunami lock, stock and two smoking barrels. In a breathtakingly unoriginal shill for cheap, low-quality translation, a press release from the think tank bleats that:
There is far too much content being created and far too few translators or money to translate it all – or even a tiny fraction of it. Independent market research firm Common Sense Advisory contends that translation automation tools such as MT promise to increase the volume and accelerate the pace of words rendered into other languages. “Translation strategies that rely on human output alone have already been overwhelmed by the explosion in content and the imperative to rapidly enter new markets,” says Don DePalma, Chief Strategy Officer at Common Sense Advisory.

About as subtle as a stomp in the face, isn’t it? Well, so much for independent analysis of the translation industry. Every time DePalma opens his mouth, a sales rep at Lionbridge gets sexually aroused. But, hey, peddling this sort of drivel as independent research gives the ideology behind cheap translation a thin coating of respectability.

At the risk of also beating a dead, decomposing horse, let me reiterate as briefly as I can the argument against the Content Tsunami: The amount of text that requires translation has not grown substantially. When we talk about a data deluge transmitted through the Internet, it involves mostly graphics and video, so the total giga and mega and googol figures breathlessly bandied about by L10N gurus are deceptive (often consciously so).

But let me rely on someone else’s words to drive home this point. In a New York Times review of James Gleick’s The Information, Geoffrey Nunberg comments on the conceptual mushiness of the term “information.” It can mean both “data,” which is meaningless, and “text,” which is meaningful (yes, even if it was written by Gertrude Stein). The slippage in and out of these two (radically different) conceptual realms is what allows charlatans to bang on about the dreaded Content Tsunami. To quote Nunberg:

When he [Gleick] describes the information explosion, he reckons the increase in bytes, citing the relentless procession of prefixes (kilo-, mega-, giga-, tera-, peta-, exa-, and now zetta-, with yotta- in the wings) that’s mirrored in the proliferation of smartphones, tablets, game consoles and windowless server farms.
 But there’s no road back from bits to meaning. For one thing, the units don’t correspond: the text of “War and Peace” takes up less disk space than a Madonna music video. Even more to the point, is “information” just whatever can be stored on silicon, paper or tape? It is if you’re Cisco or Seagate, who couldn’t care less whether the bytes they’re making provision for are encoding World of Warcraft or home videos of dancing toddlers. (Americans consume more bytes of electronic games in a year than of all other media put together, including movies, TV, print and the Internet.)
But those aren’t the sorts of things we have in mind when we worry about the growing gap between information haves and have-nots or insist that the free exchange of information is essential to a healthy democracy. Information, in the socially important sense — stuff that is storable, transferable and meaningful independent of context — is neither eternal nor ubiquitous. It was a creation of the modern media and the modern state (Walter Benjamin dated its appearance to the mid-19th century). And it accounts for just a small portion of the flood of bits in circulation.

Please allow me to reiterate the point, because it is crucial and reveals the flimsiness of the thinking by l10n gurus devoted to selling redundant MT systems: “Information, in the socially important sense — stuff that is storable, transferable and meaningful independent of context — is neither eternal nor ubiquitous… And it accounts for just a small portion of the flood of bits in circulation.” That means that truly valuable information, the type that cures cancer or leads to innovative inventions, has not grown significantly.

But don’t expect something like the truth to deter a dead-horse-pummeler as consummate as Mr. DePalma. Remember that you can’t convince someone of something if his salary depends on his not being convinced of it. This sort of implores one to ask the question as to how the Common Sense Advisory funds itself. I find it hard to believe that real companies actually pay this think tank for its “expertise” when they can get the same “World According to the Commonplace Advisory” Weltanschauung for free from the other McLocalization gurus.

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.

Friday, November 25, 2011

Transperfect Morphs into Giant Patent Troll

Bialystock (reading adaptation of Kafka’s  
Metamorphosis while looking for worst play ever written):
"Gregor Samsa awoke one morning
to discover that he had been transformed
into a giant cockroach..." (Pauses.) Nah, it's too good.
—The Producers (1968)

Do you want a measure of The Great Stagnation in Localization Land? The mosquito-infested backwater of language technology now has its very own patent troll! Yes! Transperfect’s purchase of WorldLingo has nothing to do with the acquiree’s technology (which is probably commoditized anyway). According to David Grunwald, the bold merger is a tactic by Transperfect to seize WorldLingo’s patents and then use them in ongoing litigation against a competitor in the field of localizing websites for large corporates. In other words, Transperfect is becoming a patent troll.

What is a patent troll, you ask? To phrase it in the style of Rolling Stone’s Matt Taibbi, it is the bloodcurling sound of the technology industry eating its young. The issue was brought to mainstream attention by a recent podcast from the teams of This American Life and Planet Money (When Patents Attack). (By the way, have I recommended This American Life? Did I mention it’s free?) To put it succinctly, a patent troll is a company that buys up reams of patents from other companies and then uses them to sue innovative start-ups. The objective is simply to extort money. There are entire companies in Silicon Valley that do not create anything but simply snap up legacy patents and use them as part of a tech version of racketeering (“that’s a really nice technology company you have there; it’d be a shame if anything were to happen to it…”). And guess who is the largest patent troll in the Valley? None other than Nathan Myhrvold. And who is Nathan Myhrvold? None other than one of the co-founders of Microsoft. Great Stagnation, Exhibit A: “Thy name is Myhrvold.”

What is wrong with that, you ask? Patents, after all, were invented to protect the work of the inventors who make our lives better. Well, the problem is that in the IT sphere the idea of a patent is problematic. One example is the idea of using the Internet to distribute a piece of software via downloading. The software itself, of course, should have a copyright. That is not the issue. The idea of downloading software via the Web in general, well… not so much. And American judges and the U.S. Patent Office until twenty years ago were inclined to this common-sense view of technology: The hurdles to get an IT patent were many and difficult to leap over. That is, until the tide started shifting in the 1990s. And now you basically can’t move a single inch in Silicon Valley without stepping on a the tails of companies holding reams of useless patents, whether for offensive or preemptive purposes. Indeed, one of the causes for Tyler Cowen’s innovation slowdown may well be the peculiarities of patent-trolling and the U.S. legal system.

What does that mean for the translation industry? This means that Transperfect is probably planning to use a patent that describes in vague terms a system that employs, say, machine translation, translation memory and terminology management over the Web. None of these ideas by themselves is original. You didn’t need a genius to come along and “invent” the idea of using computers to translate text. Or of recycling previous translations. Or of using the Web to harness the power of the hamster mob. But (wham!) Transperfect is going to leverage those dead patents to hit a competitor.


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.

Friday, November 18, 2011

Kumail Nanjiani on the Age of McLocalization



I first heard this comedian on the This American Life podcast. It was a short two-minute bit that closes his show called “Cheese.” It is one of the funniest things around, mainly because of the stress he places on certain words. Go ahead and watch it here.

Another clip taken from his act has Nanjiani discussing a video game called Call of Duty (full disclosure, I also play it). He remarks that previous installments of the game were based on real wars, but subsequent versions of the game have been set in current conflicts such as Afghanistan and Iraq. He expresses surprise that one level is even based in Karachi, the city in which he grew up. The joke is that he thinks this might give him an edge in playing: “Hey, guys we can hide out there. I used to rent videos there. Mr. Sadiqqi will let us hide out.”

But then he goes on to make a further observation that is symptomatic of the current state of McLocalization and how it combines a very unconscious Anglocentrism with a naïve but clumsy universalism. The game makes a huge investment in creating realistic settings. Except for one small detail, as Nanjiani points out: 

The name of the language you speak in Pakistan is Urdu. That’s the name: Urdu. But all the street signs in Karachi in Call of Duty are in Arabic. Yeah. Totally different language. And I know that doesn’t sound like a big deal, but this game took three years to make. If you look at it, they get every detail perfect, like the graphics. You can see individual hairs on people’s heads. When they run, they sweat. When they run, their shoelaces bounce. All they had to do was google “Pakistan language.” They were literally like: “What language do they speak in Pakistan?” “Uh… I don’t care. I can’t get this guy’s sideburns right.”

Watch the whole video here:



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.

Monday, November 14, 2011

Whenever I Hear Agency Owners Griping About Translators…


I keep hearing agency owners whining about how difficult it is to recruit over the Internet. “People found through ProZ still have to be vetted!” “Linguists with excellent qualifications on paper turn out to be a disappointment!” My reaction usually is: "What? The guy whose mailing address is the same as a provincial prison in Kuala Lumpur turned out to be unreliable? What is the world coming to?" "You can’t rely on people found through the same medium used by Nigerian scammers?" "Cancel that Russian mail-order bride!"

Recently, one of the “innovators” in the field of crowdsourced translation launched a “bad translators” list where you can denounce unreliable linguists (only humans need apply). (I will not link to him because: 1.- I think the dude is more than just a little insane and 2.- I suspect he is provocative on purpose to stir up a hornet’s nest and drive linkjuice.) But the question stands: Are things really that bad out there? (One observation: It is highly indicative that the people who think translation is a commodity are usually the same people who gripe about the unreliability of these translators found on the Internet.)

I think sometimes we over-adapt to new technology. We think it is better than it actually is. And then we get frustrated because it develops slowly or because it doesn’t deliver the perfect world that only exists in our heads. I think that is the problem here. Nobody stops to think that access to thousands of freelancer profiles might be a double-edged sword.

As a rule of thumb, the quality and value of a professional relationship is probably inversely proportional to the degree to which it is formed and nurtured online. For both sides of the equation. If the relation was hatched via the Internet, both the agency and the freelancer probably view each other as much more repleaceable than they would otherwise. That is just the economics of the Internet.

Secondly, the best client relationships I have had were achieved through references from colleagues and word of mouth. That is how good companies recruit. Blogging, tweeting, online networking and all the rest of that is fine and dandy. But who you know in person counts much more. The world is still very, very local. The constant propaganda about how we are instantly linked to the entire world is deceptive on many levels. We may be separated by six degrees from everyone, but those from whom we are separated by one degree still matter much more than Kevin Bacon. Ignore that at your peril.

Furthermore, I don’t know if agencies realize that the very inefficiency of the market is partly what keeps them in business. In a completely disintermediated world, the hassle of vetting translators would disappear. But in that frictionless world, agencies would be as extinct as the Tyrannosaurus. Providers and direct clients would silently drift toward each other without the messy middle man. 

So vetting translators is a hassle? I’m sorry, but if you don’t do that, what is the service that you provide, exactly?

To know where you fall in the necessary/unnecessary spectrum, ask yourself the following questions:

1.- Do you know the marital status of your freelancers?
2.- Do you know if they have kids?
3.- Can you name something about them that isn’t on their CV?

If the answer to any of these three questions is “no,” then you should probably rethink the way you have structured your business.


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, November 10, 2011

Think Different/Differently: Is Translation a Commodity?


David Grunwald acknowledges my hat tip to his Ortsbo coverage but reiterates that translation is a commodity. His basic argument is that translation does not involve creativity. I half agree with the premise that translation does not imply creativity in the same sense that artistic creation does, but I also reject the corollary. You can’t conclude from this starting point that translation is a commodity. Please allow me to highlight the fact I remain astonished by the sight of the owner of a translation agency publicly stating that he believes translation is a commodity. First of all, translation is a service. If you provide service as a commodity, how do you distinguish yourself from the competition? There are only two ways: fiercer marketing or lower rates. That is the eternal race to the bottom that leads us to the clowns from Lackuna. 

The problem with the “is translation a commodity?” debate is that no easy argument exists to prove that it isn’t a commodity. As in the case of many cultural objects —such as humor or art— you either see its value or you don’t. A joke that needs to be explained and dissected ceases to be a joke. Perhaps you can fall back on easy stereotypes about engineers. "They just don’t get why a Gauguin might be interesting despite the lack of realism." But that is a lazy straw man. The tech industry is not universally blind to the quality of outward presentation or design. The “Low Quality Translation” people are not totally tone deaf. They simply (and willfully) shut out evidence that falsifies their industry-centric world view.

If translation is a commodity, so is copywriting. But listen to Steve Jobs obsessing over the 1990s slogan that led his company to charge back to the top of the worldwide tech industry. This is an excerpt from Malcolm Gladwell’s recent review of Walter Isaacson’s biography:

The famous Apple “Think Different” campaign came from Jobs’s advertising team at TBWA\Chiat\Day. But it was Jobs who agonized over the slogan until it was right:

They debated the grammatical issue: If “different” was supposed to modify the verb “think,” it should be an adverb, as in “think differently.” But Jobs insisted that he wanted “different” to be used as a noun, as in “think victory” or “think beauty.” Also, it echoed colloquial use, as in “think big.” Jobs later explained, “We discussed whether it was correct before we ran it. It’s grammatical, if you think about what we’re trying to say. It’s not think the same, it’s think different. Think a little different, think a lot different, think different. ‘Think differently’ wouldn’t hit the meaning for me.”

This type of process is not much different from the daily process of translation. It is the laborious wordcraft that goes into writing and its sister métier of translating. So then, is translation a commodity? That is an interesting debate that is difficult to resolve. But what is certain is that there is a golden rule: Your degree of conviction that translation is a commodity is inversely correlated with your profit margins.


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.

Monday, November 7, 2011

Learn English with a Totalitarian Dictator: The Cost of Lower Quality Translation

Kathleen: It happened in Spain. People do
really stupid things in foreign countries.
Frank: Absolutely. They buy leather jackets for much
more than they're worth. But they don't fall in love with fascist dictators!
—You’ve Got Mail

There is an asymmetrical relationship between the passage of a text from English to Spanish compared to the translation from Spanish into English. And this asymmetry is frequently unacknowledged, especially by people whose background is more technical than linguistic. For example, Silicon Valley is a very diverse, multicultural, multiethnic environment, but only superficially so. In reality, it is very homogenous: It is inhabited by engineers from all parts of the world. And it is also fiercely monolingual: English is really the only language used. It is almost inevitable that the sort of translation philosophy that would come out of this inbred environment is the “Lower Quality Translation” movement. And that this philosophy would be deaf to the asymmetry I mentioned above.

The asymmetry is this: An American or a British company might opt for a bad Spanish translation as a decent stand-in for a better text (Americans and Brits rarely have a clue about what it is like to live in a world where a lot of the text consumed on a daily basis is badly translated). In stark contrast, a (self-respecting) Spanish company would never choose that option when translating its material into English because, in Spain and many other parts of the world, Pidgin English is routinely an object of ridicule. It marks you out as a rube, as unsophisticated. (And I am guessing that stigma is far more “universal” than the urge to communicate at the cheapest possible cost.)    

For empirical proof, look no further than these YouTube clips of Francisco Franco reading out a proclamation in English



I think the video comes from the early days of the 1936 military insurrection that triggered the Spanish Civil War. Someone on Franco’s staff must have decided that an English-language speech was just the trick to get the British Government and public to side with the right-wing coup. One of the YouTube clips is entitled “Learn English with the Generalissimo.” Franco read the text phonetically and none of his aides took the trouble of smoothing out the rougher passages. (The subtitled Pidgin English video is even a mini-genre in Spain; this is from a taped message by the head of a major bank communicating with his British employees).

Like Wittgenstein’s duck-rabbit, the literal meaning is sharply different according to the audience: perplexity for native English speakers, extreme hilarity for native Spanish speakers. Lower quality translation has a reputational cost that is hard to quantify in a monolingual English-speaking milieu, but which is nonetheless very, very real.

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, November 3, 2011

Ortsbo and the Sleaze Factor


Sorry it had to go down like this, baby, 'cause yo ass is cute. But you got got.
—The Wire

you got got: you've been had; you've been tricked.
Hey man, what about our deal?
Heh! You got got!
—Urban Dictionary


I have been following David Grunwald’s series of posts about a language technology company called Ortsbo with interest for some time now (read the entries here, here, here and here). He has been blogging about it for months. In a nutshell, Ortsbo is a public company that trades on the Canadian stock market and purports to have created a program that allows you to translate instant messages automatically. The company periodically rolls out announcements that it has launched an application of its proprietary technology for some use or other. The problem is, as Grunwald has demonstrated, that it is a total scam. The company does not have its own automatic translation software. It is hooked up to the Google Translate API. It is pretty straightforwardly a rip-off. Just look at the following YouTube presentation by an anonymous user:


In brief, it is a massive scam of the old “pump and dump” type familiar to anyone who has watched Wall Street.

The thing I find baffling (aside from the passivity of Canadian market regulators, especially in the wake of the Sino-Forest scandal) is that, although Grunwald’s blog is widely read, on Twitter you regularly see gurus and pundits retweeting Ortsbo’s press releases or circulating its bogus operational reports. Why is this happening? There are only two possible explanations: ignorance or connivance.

Ignorance? Well, there is plenty of that around in the world (and a healthy percentage of it in Translation Consultancy Land).

But what do I mean by connivance (please note: "connivance" is different from "complicity")? I am guessing that many of the consultants and analysts who fling Ortsbo’s criminal dung at each other like frenzied chimpanzees are probably aware of Grunwald’s exposé. But they don’t care. That’s it. They simply do not care. Which is cold, cold, cold, because real people who invested in that company will lose money. And other people will go to jail. But a lot of the people who contribute to spreading Ortsbo’s garbage are self-righteous, pious souls who think bankers’ bonuses are obscene and feel warm and fuzzy about Occupy Wall Street. Now, of course, no amount of malfeasance in an industry as tiny as translation will ever require a bailout like TARP, but my point is that financial plutocrats do not have a monopoly on the stupidity and compromised morals that led to the multi-tranche synthetic subprime CDO cubed.

The problem is that the people doling out the language tech feelgoods on the streetcorners of the World Wide Web are in the business of hype. As Web 2.0 ideologues like to say, the quality of information doesn’t matter. No, the amount of information is what matters. So they just circulate these press releases to keep the buzz going about the “vibrant” and “innovative” (cough, cough) translation tech sector. Because, without the buzz and the hype, the reality of a stagnant Florida swamp is harder to pass off as prime real estate. So they just keep on bringin’ in da funk and da noise. (Now, does anyone need further proof that there is an MT Bubble?)

Now, I disagree with Mr. Grunwald about most things. His ideas about translation as a commodity are depressing and I wouldn’t work for him unless something with a bit more dignity—such as “circus freak”—weren’t a viable career option (for whatever reason). However, he must be commended for his honesty in denouncing the Ortsbo mess. And if l10n consultants wish to wash off some of the sleaze that is beginning to spatter them, they should help him clean out this Augean stable. 

 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, several small-and-medium-sized brokerages, asset management institutions based in Spain, 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 or follow him on Twitter.