August 1st, 2015

A Few Remarks About Learning New Languages As An Adult

I was going to take A. to open gym today, but it was closed. I was going to have company this afternoon/evening, but they canceled. So: free day! When T. and R. get back from bicycling, I'll probably go for a walk, but for now, I'm here at home successfully NOT playing Farmville 2: Country Escape and I _am_ refreshing some of the Duolingo Dutch. But there's only so much of that I can do before I feel like self-distracting, so today's question is driven in part by having noticing just how badly the quality fell off in Duolingo in the later levels, and R.'s theory that happened because so few people worked through the entire course.

Question 1: How do people count and measure fluency in a second language in the US? In the EuroZone, because each state has its own (or more than one) language, and because there is a high degree of commitment to working together, the EU is largely working from a model of Everyone Learns English As A Second Language. There is widespread understanding that (a) it's easiest to get a British teacher and (b) you'd rather have an American accent. There is less understanding -- and I mean by everyone -- that European English is its own Thing. It has its own idioms, usage and vocabulary that does not exist in other versions of English (my go-to example: "social partners"). Because English is a working language across the EU, there are extensive testing metrics and credentialing systems for measuring proficiency.

If you ask the goog, how many people in the US someverbhere a second language, you get stuff like this:

Basically, you call random people up on the phone and ask them in English if they can carry on a conversation in a second language and you note down their answer. No testing. No credentialing. No asking when the last time they did so. No asking whether this was something simple like telling the guy at the burrito stand you want that sin queso or something more complex.

The goog also offers up more substantial and less cheery analysis like this:

This is an article that news organizations have been producing my entire life. Learn a language and the guv-mint will hire you to somethingorother in really sketchy places. Or at least read their news media and provide a precies in English of what you thought it said.

Entire language learning companies made their nut selling products and services promising to help the armed forces, the intelligence agencies and others try to turn people who speak some regional variant of American English into someone who can successfully navigate some other part of the world without unthinking and dangerous dependence on a local translator. I'm thinking of you, RosettaStone, and basically all of these companies are very quiet when they inevitably lose their contracts because basically, nothing works really well.

The article also says businesses will want you. Sure, they all say that. But they aren't that much better at actually hiring people who really speak a second language -- everyone wants to train people. (Which raises all kinds of fascinating questions. What's wrong with all those people who already speak the target language, anyway? Is this one of those "fit" things which is really about bias?) The usual conclusions are reached: executive function helps, but what _really_ works is how bad you want to use the new language, and whether you're willing to look like an idiot by using it.

This comment seems most relevant to my observation about Duolingo's drop off in quality:

"From knowing nothing to a little bit, (there are) huge changes in the brain," Osterhout pointed out."(From) knowing a little to knowing a lot, (it is) much more subtle."

The balance -- we should start teaching kids other languages when they are younger -- is probably true, if we really cared about our population being fluent in other languages. But the hard truth is that the languages we cared about a lot when I was a child are not the languages we care about right now. And that could well happen again. And I'm pretty sure that the _actual_ demand for being really fluent in Farsi or whatever is a lot smaller than the _actual_ demand for competency at the early levels of, say, calculus. Pretty easy to predict which of these two low demand educational attainments is gonna win, given that calculus is written into the prereqs for a lot of highly desirable higher education credentials. And Farsi ... is not.

I'm ignoring that guy over at Fluent in 3 Months because ... he's That Guy Over At Fluent in 3 Months.

"Obama is absolutely right. We live in a global economy and whether it's Spanish, French, Russian or Chinese, Americans lag behind the rest of the industrialized world when it comes to mastering a second -- or even a third -- language."

Notice that the author identifies as an "international careers expert". So _of course_ she's gonna say ya gotta do this. This is what she sells. As for lagging behind the rest of the industrialized world, well, there's a reason for that. They're all busy learning our language as a second language. If your first language anywhere in the world is Not English, the next obvious language to learn is ... English. With an American accent. It's a lot less obvious what the next language to learn is if you grew up speaking English with an American accent. (<-- Does this sound imperialist to you? It sure as fuck should!). The balance of the article is an ad for Praxis, and arguments in favor of learning Spanish for travel to, say, Pamplona for the Running of the Bulls. Or Chinese, the language of global commerce. (<-- I'm quoting.)

This high school Spanish teacher is death on grammar based school lessons, a big believer in the use-it-or-lose-it principle, and a hardass when it comes to immersion: she married a native speaker of Spanish and hired a Spanish speaking nanny for the kids. She says: "This component to your studies is absolutely necessary--you MUST experience immersion. Studying or working abroad after having gained a good base is what it truly takes if you want to become fluent in a language. From my own experiences abroad, I can assure you that there is no substitute for having to navigate your daily routine and maintain personal relationships in a language that is not your own. Students who take four to six years of language in a classroom setting find their proficiency level high enough that they can reach fluency in a matter of months abroad."

If this is the criteria for Americans to attain fluency in a second language, I think it is safe to say that This Ain't Gonna Happen. For one thing, the same guv-mint that supposedly wants to hire fluent speakers of languages spoken in places where Americans Aren't Real Popular Currently At Least Not With the Existing Regime, that guv-mint absolutely will not hire anyone _from_ those countries, or with too many connections or time-in-country, in those countries. Loyalty reasons. I so did not believe this at first, because I was like, what? But you know, Empire (I'm not talking about the TV show here).

This blog entry at the Economist is critiquing the idiots over at Freakonomics:

Of course, all of the arguments made in this gilded turd apply way more strongly to studying computer science -- and the payoff for get a CS degree are widely understood, yet we still have a shortage, which suggests that this may not be a motivational issue.

Bizarrely, the author tosses this nugget out:

"What is the return on investment for history, literature or art?"

Wow. If you want to get foreign language more space in the school system, you'd better learn to work _with_ the other people already there, not against them. Also, I would argue at least in favor of learning history, simply because the personal payoff (and I mean dollars in net worth) has been so ludicrously huge.

The NYT weighs in with a new frame. Maybe we _aren't_ really a monolingual country.

Great quote from Arne Duncan here, proving, as always, that people who attain the highest, most rarefied levels of education are astonishingly foolish:

"“For too long, Americans have relied on other countries to speak our language,” Secretary of Education Arne Duncan said at the Foreign Language Summit in 2010. “But we won’t be able to do that in the increasingly complex and interconnected world.”"

Er, we won't? Interesting assertion. I'd love to hear the back story on that one, because the current evidence suggests that it's just gonna get easier as we go along. I mean, if the EU had settled on something _other_ than English for their working language, sure, I could see the argument. But now? Yeah, right.

Anyway. In this article, the author compares the census question (Anything other than English spoken at home?) and says that's the wrong question for assessing bilingualism in the US. He compares it to the EU question (can you have a convo in some other language returns!) and says the EU isn't as multilingual as we assume, and the US isn't as bad off as we assume. He may well be right -- I've certainly found people in the Netherlands who speak little to no English, and we got through on my Dutch (for the record, the main one I remember was a truck driver in 2002, so don't be thinking that failure to speak English is a route to great success; the rest were restaurant workers in Ameland, where we got by primarily in Dutch, because their two languages didn't include strong English).

This article is worth reading. The author has written a book about people who learn languages. I'm going to go off to read reviews of that at Amazon and decide whether it's worth reading.