Don't TEACH a foreign language...


The traditional way of teaching foreign languages makes no sense. There is such thing as human nature; schools that ignore it pay a price. The big idea of our school-in-the-works is that if we can understand human nature — and if we can re-create education to harness it — we can build a better method of schooling than anyone has ever seen before.

Maybe nowhere is the silliness of traditional schooling better seen than in the way foreign language has been taught. The default model of teaching foreign languages doesn't work. In fact, given what we know about human nature, it can't work.

What I'll dub the "traditional model" of teaching foreign languages has at least two egregious faults —

  1. Foreign language instruction begins in adolescence, after the critical period for language learning has closed.
  2. Foreign language is learned 'academically': as alien words and rules to be committed to memory. That is, language is learned like any other academic subject — e.g. astronomy, or trigonometry, or anatomy.

If we want to bring up adults who are fluent in multiple languages, the traditional model is nonsensical. But there is a natural way to learn foreign languages, and our school can harness it.

The tricky word "natural"

Let me suggest something that will work against my overall point: mistrust anyone who talks about "natural" ways to learn. This includes me, and anyone else speaking for our school!

Truly natural skills don't need to be taught: say, walking, pooping, and talking. The purpose of schools is, on the contrary, to teach unnatural skills.

Reading is unnatural. Math? Unnatural. Science, too! And so on.

Academic subjects aren't natural: that's why we teach them in school.

I realize that these are fightin' words, and indeed I am overstating this point. Really, the natural/unnatural divide is full of problems, not the least of which is that it's the nature of Homo sapiens to do artificial things! It's natural for us to do the unnatural.

And in fact even the most artificial academic skills (like reading) are eked out of our deep nature. Steven Pinker wonderfully quipped, "a group of children is no more likely to invent an alphabet than it is to invent the internal combustion engine," but of course both an internal combustion engine and an alphabet were initially created using our basic cognitive attributes — our skills and our motivations. Otherwise we couldn't have created them at all.

(The fight Catherine and I had in the comment boxes about math learning some months ago demonstrates the silliness of arguing whether something is natural or unnatural. Catherine argued math was natural, because humans do pattern recognition. I argued math was unnatural, because we don't do complex, abstract, quantitative pattern manipulation. We were right.)

But I'm overstating this point to make a larger one: language learning is an exception to this rule. Learning to speak and understand a language is wholly natural. Talking is in the same camp as walking and pooping! Every baby learns to master a foreign language, and they do it without the incentive of grades, and without the intervention of textbooks or intentional teaching at all.

There really is a purely natural way to get kids to learn a foreign language, and that's to merely re-create what already goes on in childhood: enmesh kids, when young, in a community using the language.

That is, the teacher shouldn't say, "Now I'm going to teach you a word that means 'cat'. It's this: 'gato'!" Rather, the teacher should pick up the gato, and say, "Ooh, este es un gato!" (And maybe: "Le gustaría acariciar al gatito bonito!")

Immerse kids in the language. Use language to help do things together. Their brains — their wonderful, human brains — are already equipped to do the rest.

Don't teach a foreign language — teach in a foreign language

All right: how can we do this?

The millennia-tested way to learn a language is to be fully immersed, as a child, in the language. That's outside our scope — it truly takes a village to accomplish this. (A village which the kids only rarely leave.)

The modern academic version of this, of course, is to do full-school-day immersion. And, indeed, the evidence shows plainly that full-immersion classrooms work. But this, too, is outside our scope — at least as we start the first of these schools. We're going to have our hands full finding teachers of a high enough skill set — we can't afford to add "fluency in a foreign language of our choosing" as a hiring requirement.

More recently, other schools have done half-day immersion. The evidence seems to be that they work quite well, too. Again, however, this seems outside our capacities as we start our flagship school.

Eventually, it would be wonderful if we could do full-day immersion. In fact, it might be even more wonderful if we could do dual-half-day immersion, in two foreign languages (say, Mandarin and Spanish, or French and Japanese). There doesn't seem to be an upper limit to how many languages a single mind can master — it would be wonderful if our schools could push the envelope here (or at least the American envelope)!

There's another option, however, between half-day immersion and the bound-to-fail "teach the language explicitly" method: content-based instruction. In content-based instruction, individual academic classes are taught in a foreign language. That is, they don't teach a foreign language — they teach in a foreign language.

The breakthrough moment for me was realizing — thanks to the suggestion of my friend L., a fantastic Mandarin teacher — that we could do this with as little as one hour each day.

And it turns out that some schools are already doing this!

Which course (or courses) should be taught in a foreign language? In principle, any of them could be. In practice, I'll speculate, some classes are better candidates than others — especially those that deal directly with the physical world, where verbs can be acted out and nouns can be picked up.

Our drawing/art class strikes me as a good choice for this. So too our physical education class.

For a different reason, math class strikes me as a potentially good option — the verbiage here is more restricted. John Mighton, creator of the JUMP math curriculum, pointed out in a webinar last week:

Eventually, we want kids to be able to explain what they're doing in words. But at first, that can be distracting. Initially, you can get at the big ideas better with sparse language.

Or, perhaps a combination of multiple classes. We'll need to think more about this as we progress, and look to the specific skill sets of some of our teachers.

What an exciting idea — that all of our students can achieve some level of real skill in a foreign language. And we can accomplish it not by teaching harder, but simply by re-jiggering the curriculum to comport with human nature.

I'll be posting a few more thoughts about this on Monday — if you've ideas, questions, or thoughts, please send them to me!