How do and from what/whom do children learn language? My question relates specifically to houses that may be bilingual and/or parents who would like their children to learn more than one language. Is there an advantage or disadvantage to overall language development if a child is learning 2-3 languages at once? Does a child learn exclusively from human interaction or can a child whose parents speak English, learn Spanish from watching Spanish cartoons, etc.?

This is one of the most long-standing debates in human history, so I may not be able to provide a definitive answer. Some predisposition to language learning is thought to be inborn, but teaching is also necessary. (A deaf child not taught sign language will invent some gestural communication, but not as sophisticated a system as the one he can be taught.)

There is an interesting researcher Patricia Kuhl (Scientist in the Crib, by Gopnik & Kuhl) who studies development of perception. She shows that although babies born everywhere have the ability to perceive any language sound ("universal phonemic library"), they gradually over the first year or two begin to prune their perception to fit what they are hearing. (eg the l/r distinction for Japanese speakers, maybe the v/b for Spanish) She has proven that interactive play in a second language maintains the ability to hear the phonemes of that language after the mono-lingual kids have lost the ability. She has specifically demonstrated that videos of these same play interactions do NOT have any effect.

Hearing phonemes is not quite the same as learning a language, of course, but it implies that beginning exposure in the first year is effective.

Children who have normal ability in the language learning part of their brains seem to have the ability to learn several languages early simultaneously. Most writers say it works best if you start before three, but some kids are more talented at it than others, and lots of kids become completely fluent after that age. Usually people who learn a second language after puberty retain an accent no matter how good their mastery of the language is. Kids with language disabilities eg autism, MR, auditory processing disorders, etc, struggle with learning any language, so most people in the ed field recommend picking one language for therapy, home and school, but of course you don't know who they are until they're toddlers. I've seen a few kids who definitely don't fit this rule, so it may not always be right, although it makes sense.

Language ability is genetic to some degree. Language development correlates well with maternal education. Contingent interactive play with eye contact is the best way to teach language-it is embedded in social interaction and emotional meaning. One to one interaction with adults who really care what you're trying to communicate is more useful than eg starting day care to be with other kids. Play with other kids has a lot of advantages, but if they don't understand you they'll just push you down.

There are some hair-raising studies from a few years ago showing big gaps by SES in average vocabulary, average number of words heard by kids, kind of language used (directive vs descriptive) which probably have a big effect on language learning. This has a lot to do with the Reach out and Read idea, which is not just to teach words, but to promote language interaction between parents and kids.

When a kid is delayed in early language most parents push for word production. Since receptive language precedes, and is necessary for, expressive language, it makes more sense to increase what's going in as well. That's why reading to kids is so useful at every level.

Multi-lingualism in a normal child should not impede language learning. Whether it enhances it is an interesting question. Maybe it enriches it long-term, but I don't think there's any evidence that it changes it short term. Except maybe sign language. It's safe to say we should not attribute delayed language in our clinic patients to bi-lingualism.

Recently sign language has come to be viewed as another language, learned in more or less the same way as spoken languages. There are some early differences, though, in which signs for simple objects and commands can be learned a little earlier or more easily by some kinds of brains. Some retarded and/or autistic kids can learn to sign a little even if they don't talk.

There's a fad among educated parents to teach sign language to babies early with the idea that they will communicate earlier and it may give them a better start in spoken language. Evidently they do learn to sign a few "words" before they speak them. It is true that everything you learn can function as a template for anything else, and this idea certainly applies to language learning. One commercially available program in NYC is called Baby Signs, and they have classes and tapes to use. I think it's a cute idea and certainly promotes a nice kind of interaction. I couldn't say if it makes any difference for normal kids. It is interesting to speculate if it would help or confuse all these language delayed toddlers we send to EI.