Best time to learn a foreign language: between birth and age seven?

Some scientists seem to think that that's what new research shows (hat tip: Steve Kaufmann at The Linguist on Language), but I'm not fully convinced because I don't think the scientists have ruled out that the cause of young children's language-learning ability is the kind of exposure they are getting rather than some innate age-linked ability.

Just to scratch the surface, let me point out some of the advantages of small children's learning methods.

Read more... Little kids get to learn solely or primarily through real-life exposure. Complete immersion in a language is not needed, but lots of exposure is. I can take my own four-year-old daughter as a case in point. She speaks English, Japanese, and Chinese, which we've pulled off mainly by me speaking only English with her, my Japanese wife speaking only Japanese with her, and hiring only Chinese-speaking caregivers who of course only speak Chinese with her. This exposure has gotten her to a point where she has been able to swing classes with native-speaking peers in all three languages. Could you imagine sitting down kids seven or under and going through a textbook with them and getting the same result? Ain't gonna happen, so kids are stuck (actually, I should probably say blessed) with learning through exposure.

The second thing is that the exposure is meaningful to them; if they want some juice, they'd better learn to ask for it. The same is not true of most of the textbook-based teaching methods that schools use. A lesson on buying shoes in Paris in a French textbook probably won't have any immediate applicability for you. Actual exposure will stick in your brain much better than constructed exposure, and actual exposure is all that kids under seven will typically get.

The new research also points to another benefit seems to be pretty clearly a method differentiation that benefits young kids:

Recall that Japanese "L" and "R" difficulty? Kuhl and scientists at Tokyo Denki University and the University of Minnesota helped develop a computer language program that pictures people speaking in "motherese," the slow exaggeration of sounds that parents use with babies.

Japanese college students who'd had little exposure to spoken English underwent 12 sessions listening to exaggerated "Ls" and "Rs" while watching the computerized instructor's face pronounce English words. Brain scans — a hair dryer-looking device called MEG, for magnetoencephalography — that measure millisecond-by-millisecond activity showed the students could better distinguish between those alien English sounds. And they pronounced them better, too, the team reported in the journal NeuroImage.

So someone saying a word slowly and clearly to you numerous times helps you get the pronunciation down. Shocking!

And the list goes on. Kids often get corrected by the adults around them, something that can only be done on a much more limited basis in a classroom setting. Kids aren't doing much else beyond learning languages, so they're not limited to 45 minutes a day of classroom exposure to the language. Etc., etc.

Now take people eight years old and up. How do they typically learn? The old-school classroom method. Actual exposure trumps textbook/lecture exposure any day, so it's hardly surprising that little kids are at such an advantage.

So, while there are certainly some biological aspects to small children's language-learning ability (small infants' ability to distinguish sounds, lost by the time they're one year old, comes to mind), I'd be reluctant to conclude that a large percentage of children's language-learning ability comes from some biological advantage when their learning methods' advantages seem so obvious.