Friday, January 20, 2012

The Undercover Economist Follows Up My Podcast Suggestion

As you know, my religion is skepticism. I flatter myself that even in front of the pearly gates, I would be withholding judgment up until the moment I am really, really ushered into the eternal hereafter of bliss, seventy-two virgins and all the pepperoni pizza you can eat. I was not educated as a scientist, but I have a healthy admiration for science as a type of controlled skepticism. Now, for the practicing skeptic, there is no better brain fodder than Tim Harford’s More or Less podcast on the BBC. Harford is perhaps best known as the author of the highly recommendable Undercover Economist, one of the pop economy books for general audiences that have become so popular in the past decade and which spawned the column on the Financial Times of the same name. The More or Less program is not directly concerned with economics; rather, it mainly consists of minute analyses of statistical data that generally conclude that most of the figures bandied about in the media and by pundits are pretty much rubbish. (If you are into finding out what drives skeptics to be so caustic about the media in general, I can’t resist plugging Guardian columnist Ben Goldacre’s Bad Science: Quacks, Hacks, and Big Pharma Flacks).

Now, some weeks ago, President Hugo Chávez of Venezuela stated that the fact that he, President Lugo of Paraguay, former president Luiz Inácio “Lula” Da Silva of Brazil, current President Dilma Roussef of Brazil, and President Cristina Fernández de Kirchner of Argentina all got cancer was evidence of (wait for it) a secret CIA plot. Although Chávez has accustomed world media to batshit claims, this was outrageous enough to warrant some passing coverage. (A friend of mine joked that maybe Chávez will make his next few appearances covered in aluminum foil to ward off the CIA’s top secret cancer ray). Anyhoo, the incident is indicative of the distortion of reality and reason that is a feature of daily life under the Bolivarian Revolution. A statement that would have previously led people to question the mental health of the head of state is just a weird little footnote to fill up some air time. Another feature of the statement is that it is typical of the extreme paranoia that characterizes the Latin American left: the CIA as an all-powerful, hidden demiurge that controls everything. This creates the same problems that the concept of evil creates for the idea of an omnipotent God. If the U.S. spy agency is all-powerful, how did it allow episodes such as the Mexican Revolution, the Cuban Revolution, Chávez, Morales, Lugo, Juan Domingo Perón, etc., etc., etc.? It reminds me of a line in De Niro’s Good Shepherd, where one character asks why you never place a definite article before CIA, and another one responds: "Do you put the word 'the' before God?" (Back in the fifties, the acronym CIA was not preceded by a definite article, contrary to the case now.)

Therefore, a bunch of aging leftist Latin Americans get cancer and the best hypothesis is that the CIA has a cancer ray. William of Ockham must be rolling in his grave. Not being a statistician, I am unable to work out the probability. So I wrote to Harford and suggested that this claim was right up the alley of the More or Less team. He agreed and devoted the first part of this week’s podcast to the issue. And here it is:

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.

Wednesday, January 11, 2012

$#*! My Localization Guru Says (‘Crowd’ Translation Means Nothing)

Do words mean whatever you want them to? That’s what Humpty Dumpty claims in Through the Looking Glass and it seems that the localization pundits who pop up in every single translationl10n conference also believe this. And when called on it, they fall back on half-digested relativistic mumbo-jumbo about sociolinguistics. Who knows? Perhaps they’re right. Perhaps their words can mean anything they want them to. So, instead of listening to a localization pundit, maybe it is more useful (and, BTW, cheaper) to stay home and read the instructions on the back of an aspirin bottle to learn about internationalization of business processes in a multilingual context. Because if words mean anything, the aspirin bottle can tell you how to manage large-scale multilingual projects.

To illustrate this little piece of wisdom, let me tell you a little story (identities have been concealed to protect the guilty):

A l10n guru tweets this:
"Crowd" xl8 of a medieval mss (link)

Like the fair-minded skeptic I am, I click on to the article in question. It is from the Chronicle of Higher Education. A 12th century manuscript was discovered by chance in an American university library. The professor decided to involve her graduate students and a few undergrads in the process of transcribing, describing and translating the text. Here is the description on the project’s homepage:

In January 2011, Professor Marie-Claire Beaulieu and her Medieval Latin students, in collaboration with Alexander May and Christopher Barbour, set out to learn more about the Tisch Miscellany. Each student chose to work on a leaf according to his/her personal interests. As each student progressed in learning to decipher the hand or print of their leaf, interesting and surprising discoveries were made… These discoveries required patience and hard work. Since each leaf is unique in the collection, each student had to get used to a different set of idiosyncrasies such as inconsistent abbreviations, non-standard letter-forms, and scribal errors. Furthermore, working with unedited or unknown texts implies complete self-reliance on the part of the editor. Students had to make educated guesses to resolve abbreviations, decipher difficult hands, or fill lacunae in their leaves without referring to other scholars’ work.

So far, so good. But note this: nary a mention of a “crowd” so far. Incidentally, a lot of academic research is actually done by graduate students. Do you think an academic showman like historian Niall Ferguson actually writes the mega-bricks he publishes under his own name? Translator, please. Ferguson is the Stephen King of academic history. He should really be called “the committee called Niall Ferguson.” Just in the past three years he has regurgitated three 500-plus-page behemoths on Sigmund Warburg and far broader topics, such as why the West dominated the world, not to mention a little tome on the history of world finance (an adulatory biography of Henry Kissinger is due to come out this year). He also flits around the globe teaching everyone why the budget deficit will sink the American Empire, pops up on Bloomberg with alarming frequency to flash his perfect coiffure, AND, in his spare time, hosts documentary miniseries for British television. Again: please. Look in his acknowledgements to see the underpaid Harvard grad students who never see the light of day while they scribble the professor’s latest magnum opus by candlelight. Niall Ferguson himself is an instance of “crowd” scholarship.

But I digress. The author of the Chronicle piece interviews Professor Beaulieu and breathlessly inserts the buzzword: "Do you think such "crowd" translation, in which the participants aren't experts but just students interested in the subject, is the future in your field?" The professor being interviewed pushes back a little, resisting the imposition of the buzzword:

Someone doesn't need to be an absolute expert on a manuscript to translate it. If you give an undergraduate the proper tools, they can do a fantastic job. This is something my colleagues in the sciences have realized for a very long time. Undergraduates frequently work in labs on major projects.

Now notice that the professor says that the future is that undergraduates can work on graduate projects. But that is a far cry from saying that we are going to learn oodles about the Middle Ages once we mobilize a mob of slacked-jawed yokels to pore over books of hours and Carolingian chronicles. No, perhaps those kids aren’t absolute experts, but they do need a smattering of slightly arcane knowledge (uhm, ecclesiastical Latin and paleography) to carry the process through successfully. Let me stress this point: Getting undergraduates who have studied ancient languages and codicology isn’t exactly the same thing as posting digital images of a manuscript on the library’s website and asking every unemployed Joe Schmoe with two semesters of community college to provide his two cents on an 800-year-old text. This is hardly the crowdsourcing Nirvana where you make a cattle call and thousands of amateurs comb over British parliamentarian’s expenses, Wikileaks’s document dumps or Sarah Palin’s emails.

No, this was fifteen handpicked undergrads and grad students who were chosen by the professor to work in a closely supervised environment on a highly technical project. This use of “crowd” is a clear instance of how buzzwords become so omnipresent that they cease to have any meaning at all. They become ethereal fluff. Like oxygen, we never notice it because it is so abundant. If a word means anything, it means nothing. File it under “synergy,” “proactivity” and a thousand other business memes collected by Unsuck It. Perhaps that will teach your mind to drift off in the middle of an l10n conference when you hear this buzzword, because it has the same function as the “ohm” in transcendental meditation: a meaningless incantation designed to set your conscious mind at rest and dull your critical faculties.

Being the skeptical gadfly that I am, I called out the pundit on her acritical retweeting of pablum:

Me: The headline has nothing to do with the text. They are often chosen so they will be RTd by people who don't read the article.
Pundit: I'm happy to let them define "crowd" as they wish for their context -- it's an interesting exercise
Me: Interesting, perhaps, but not very valid from an intellectual standpoint. Meaningless terms beget meaningless thoughts.
Pundit: Most sociolinguists would beg to differ. People get to choose how they use language. That's how it evolves.
Me: Lazy relativism. Only the insane give any meaning they want to words. Or poets. Meaning is defined by rule-based convention.
Pundit: We owe a lot of words to linguistic creativity. Thanks, Shakespeare! The people trump language purism every time.

(A language purist? Me? By the way, if you're reading this and wondering when exactly I decided to kick your ass, it was at "language purist".) In summary, a classic case of terminological sloppiness being concealed behind a tornado of relativistic mumbo-jumbo. I leave you with another, more illustrious, case:

'There's glory for you!'

'I don't know what you mean by "glory",' Alice said.

Humpty Dumpty smiled contemptuously. 'Of course you don't — till I tell you. I meant "there's a nice knock-down argument for you!"'

'But "glory" doesn't mean "a nice knock-down argument",' Alice objected.

'When I use a word,' Humpty Dumpty said, in rather a scornful tone, 'it means just what I choose it to mean — neither more nor less.'

'The question is,' said Alice, 'whether you can make words mean so many different things.'

'The question is,' said Humpty Dumpty, 'which is to be master — that's all.'

Alice was too much puzzled to say anything; so after a minute Humpty Dumpty began again. 'They've a temper, some of them — particularly verbs: they're the proudest — adjectives you can do anything with, but not verbs — however, I can manage the whole lot of them! Impenetrability! That's what I say!'

'Would you tell me please,' said Alice, 'what that means?'

'Now you talk like a reasonable child,' said Humpty Dumpty, looking very much pleased. 'I meant by "impenetrability" that we've had enough of that subject, and it would be just as well if you'd mention what you mean to do next, as I suppose you don't mean to stop here all the rest of your life.'

'That's a great deal to make one word mean,' Alice said in a thoughtful tone.
(Through the Looking Glass, Chapter 6)
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, January 2, 2012

The Wisdom/Madness of Crowds: Obama, Hillary, New Hampshire, and Me

One of the “pop economy” ideas now in vogue is the wisdom of crowds. It is often invoked as the theoretical foundation of crowdsourcing. The idea in its current incarnation is not all that new and has some very respectable origins in economic theory. It can go as far back as Adam Smith’s invisible hand. Much of modern economics is fascinated with complex (social, economic, online) systems that function without centralized organization. Capitalism is one such system… up to a point (leaderless movements such as Occupy Wall Street or the Spanish indignados ironically imitate markets in their attempts to create non-hierarchical modes of organization). Milton Friedman’s classic example of the pencil is one point in case: thousands of people spread all over the world collaborate to obtain the wood, graphite and rubber to make the humble pencil, but none of these people know each other and no unified, central authority coordinates their work. There are no über-project managers coordinating all this movement to and fro of goods and services. No, the massive organizations that work together to build a simple pencil do so in the absence of a central planner.

One recent application of this idea is the fad for prediction markets. You can go to InTrade and make a bet on a mesmerizing variety of world events. Do you think Ahmadinejad will be ousted as Iranian president by December 31? You can make that bet there. Do you think Sarah Palin will jump into the Republican race? Ditto. It is a corollary of the efficient markets theory. According to this very influential current of thinking, one single individual cannot know what will happen in the future. But if a market is large enough and liquid enough and fair enough, those markets (which, remember, are the aggregation of choices made by thousands or millions of investors) are powerful indicators of the future. The same principle underlies our interpretation of bond markets: the variation of U.S. Treasury yields can be taken as a proxy for what investors think about the medium-term future of the economy.

I became a punter at InTrade during the 2008 U.S. presidential elections. And I had a really bizarre experience that taught me a lot about the value of the wisdom of crowds.

I was determined just to make small bets (and InTrade does allow you to make tiny, tiny bets because trading commissions are very low). I can’t have wagered more than $30 or $40 in tiny bets of a dollar or so. I don’t remember exactly what my strategy was: whether betting against conventional wisdom or making small bets on unlikely outcomes. But it was something along those lines. Anyway, in January 2008, as primary season (four years ago exactly) was revving up, I probably had ten or so bets on the first few Democratic and Republican primary races. Then I just lost interest. In InTrade. Not the election itself.

First of all, if you recall, Obama won the first primary in Iowa. Against all odds, a black man won in Iowa! I distinctly remember his victory speech. I had heard all the buzz about Obama, but I had never actually seen him speak. Then he got up on that podium. His posture was strange. He didn’t look straight at the audience. He just sort of looked askance at some point in front of him. He wasn’t at all euphoric. He was incredibly calm. He lifted up his hand and said, with that intonation: “They said this day would never come…” I literally felt chills. I had never seen anything like that. I had read about similar speeches. I had seen the tape of Bobby Kennedy speaking on the night Martin Luther King was shot. I have read Lincoln’s speeches. I read a book on Miguel de Unamuno’s “venceréis, pero no convenceréis”. But to actually see a performance like that live was totally different.

Anyway, Obama’s victory in Iowa probably blew all my bets to hell. I just lost interest in the InTrade game and got swept up in the media-driven soap opera of the campaign. The conventional wisdom that the race was Hillary’s to lose immediately shifted after Obama’s upset. Whereas, before Iowa, Hillary only needed to tap dance her way to the nomination, she was now written off. Completely. Obama was crowned as the certain Democratic candidate. Hillary was yesterday’s news. Maybe you don’t remember the events of that amazing week. I remember checking my InTrade markets. Effectively, the market was now giving Obama an above 90% chance of winning in New Hampshire, the second primary, usually held one week after Iowa.

The thing is I got bored with my bets and did not change them. And here is the key detail: I had a bet on Hillary winning in New Hampshire. Not that I actually thought she would win. It was just that I was so uninterested in the game, I left an old bet on. And then something amazing happened. You must remember it. Hillary Clinton made a campaign stop in which she sat down to speak to a group of women in New Hampshire. It was the usual political rhetoric, but the strain was visible. Clinton made remarks about what a mistake the electorate was making by choosing the wrong person (many pundits honestly thought Obama was unelectable). And then someone asked her how she was holding up. And then the most amazing thing happened. Her voice cracked. “It’s hard,” she managed to say and then her eyes welled up. The world watched in wonderment: “Oh my God, she was showing some real emotion! Hillary Clinton is a human being!” I can’t say I ever liked her, but it was a real moment. You felt all that she was feeling. All that work, all those years of being the perfect student, the Stepford wife, the good daughter, all the humiliations that led her to that moment and it was slipping away, right when it was in the palm of her hands. How could you not feel some empathy?

So it was a strange, touching TV moment. Even memorable, perhaps, but Mrs. Clinton was still heading for a crushing defeat. All the polls put her well behind Obama. And then something even more amazing happened. I remember following the CNN coverage via the Internet and simultaneously watching my InTrade positions. The first returns, oddly, had Clinton winning by a small margin. Not a strange thing with small percentages and early returns, as any political junkie can tell you. But the InTrade market moved. It trembled the tiniest little bit. Obama’s odds of winning went down from about 90% to about 85%. Hillary was still cruising for a bruising. And, since my bet on Hillary was still on, I started to make a little money on my one-dollar bet.

I won’t drag it out. As return after return rolled in, Hillary Clinton achieved a lead and maintained it against her younger and flashier opponent. Although I liked Obama, it was delicious watching the position on the InTrade screens. With every new percentage announced by the networks, the Obama and Clinton lines got closer and closer and closer, until it was 50-50, and after 60% of the returns were in, the roles were completely inverted. Hillary Clinton was being given a 95% chance to win New Hampshire and Obama only 5%: with 92% of the votes already counted.

My one-dollar bet was now worth about $14. A one time 1300% gain is nothing to sneeze at. And if you’ve ever had one, it feels a little like winning the World Cup. I know now why some people get hooked on gambling. And why traders can get hooked on what they do. Fortunately or unfortunately, I don’t have those genes. I’m not dumb. I recognized what had happened. My indolence in maintaining my old bets had left me exposed to a positive black swan, a nearly impossible chance occurrence which, oddly, happens more often than our misguided theories about statistical probability and the wisdom of crowds would suggest. I simply pocketed my ill-gotten gains and exited the market. Never to return again, nursing an (overall) ten-dollar loss (more or less) and a little bit more (individual) wisdom.

So you ask me if markets are rational? If they can predict the future? If crowds are better at telling the future than an individual?

Well, the InTrade market did predict that Hillary Clinton would win New Hampshire… about two hours after she actually won the state. So there you go.

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.