rationalskepticism.org

Mr.Samsa wrote:
Hmm.. grammar obviously does rely on concept formation,

Uh, yes. What I was getting at (very ineffectively ) was that the ability for symbolic thought comes first, and provides the basis for the ability to string symbols together into regular patterns (grammar). You could have the first without the second, but not vice versa.

but I would imagine that it's a particular instance of it that would be affected by a critical period, rather than concept formation in general. I think grammar (and phonemes, or language in general) is so complex that you have to learn the basics to the extent that they become concrete in your head and you can start to understand the rest of language from there. So as soon as you learn the fundamental phonemes of your mother language, the rest becomes a whole lot easier to learn, and obviously the more you learn then the more cemented the phonemes become.

That would be about right. In fact, I'd argue that after you've learned the basics of one language, this helps you with any other language, because you are already able to "do language". So you're able to learn a completely different language even without a mediator (or else where did Cortes' interpreters come from?), starting from the scratch of nonlinguistic clues all over again like a baby, something feral children seem unable to do.

katja z wrote:Of course, this question is only relevant if the critical period hypothesis does, indeed, hold. But as far as I'm aware this is based on feral children's problems with language acquisition, and I don't think in these cases you can separate the effects of language deprivation from the overall effects of the lack of interaction with other humans on general cognitive development.

That's a good point about feral children, but I think that if a critical period doesn't exist in the traditional sense, then there is some other process which operates in a similar way as we need to explain why people who learn one language misinterpret the sounds of phonemes from other languages. For example, I think those who learn Asian languages tend to hear the letter "L" as an "R" sound.

I was speaking about the validity of the critical periods for other aspects of language. In the case of phonemes, we know what and how it happens. It isn't an irreversible process though. It takes a lot of practice but it is perfectly possible to learn to hear new distinctions - and once you hear them, to produce them correctly enough to be functional, even though you'll probably realise the individual phonemes slightly differently from native speakers. Interestingly, there are vast individual differences here, with some people just having "the ear" for languages - I don't know what the reason is.

ETA: Just a thought, what about the critical period for sign language in deaf people? Has any work been done on that? I'll do some googling tomorrow, here in Mitteleuropa it's so late that it's already early ...

katja z wrote:
Indeed, I get that but what I meant by grammatical similarities was that whilst phonemes seem to be fairly language-specific, grammatical structures seem to be present in much broader groups, so spotting gramenes ( ) might be harder (or impossible?) as they won't be as obvious or as salient as difficulties understanding phonemes from other languages.

Hmm, it is with phonemes just as with grammatical features, some are widely distributed over a vast number of languages and some are more restricted to groups or even single languages (?). On the other hand, it is true that the phonetics of a language is probably the aspect that can change most rapidly (and at some point phonetic change will translate into a change in the phonemic system) so very often even two closely-related languages will be very similar in vocabulary and grammar but fairly different phonetically.

katja z wrote:

I imagine that the McGurk effect would have something to do with what we learn during this time too.

Can you explain? I don't understand and my telepathy module seems to be broken

Sorry I thought you commented in the McGurk thread?

(snip)

My point was simply that during the time we're learning phonemes, and/or gramenes, we probably also learn these visual cues which permanently alter how we hear sounds and learn language.

Thanks. I did comment there, I just didn't understand in what sense your remark was meant (that it was learnt during this time, or that it influenced learning, or ...).