[Corpora-List] Language Learning (was: 'Quantitative Corpus Linguistics withR'--re Louw's endorsement)

[Mike Maxwell]

Alexandre Rafalovitch and I had originally off-lined this, but since
it's out in the open, here's some back and forth between us.  I've added 
a few comments in my own parts in [].  Alexandre first (quoting me, so 
who's on first?):
> Mike Maxwell wrote:
>> My favorite brute fact (associated at least with syntax and
>> morphology-- I'll make no claims about semantics) is that every
>> child, even those with mild retardation, learns a language without
>> being taught.
> 
> How can this be a brute fact when every day you can see a  parent 
> constantly correcting child on how to say things either by directly 
> pointing the mistakes out or by repeating or rephrasing the same 
> things in the correct and accepted way. That's teaching in my books, 
> just not classroom style teaching.

There's an extensive literature on this.  Basically, the parents' 
corrections are usually beside the point, or ignored, or misunderstood. 
Probably the only place such corrections work is where there are
morphologically irregular forms.  [And in any case, apart from rote 
learning in such cases--which is what irregular forms are--the child has 
to abstract a generalization.  In teaching, in contrast, one is usually 
explicitly taught the generalizations: "The algorithm for computing the 
roots of a quadratic function is..."]

In addition, there are plenty of things that children learn about
language that not even linguists understood until recently.  Things like
the complementizer-trace phenomena, or island constraints.

> Also, at least for the Russian language, syntax and morphology  is 
> drilled into the children heads during multiple years of schooling. 
> And you can tell the person who did not undertake that schooling and 
> just learned as much as possible through absorption.

Sure, but the schooling is just reinforcing a more archaic [I should 
have said standardized] form of the language.  The unschooled person 
speaks a language just fine; it's just not the prestige variety, rather 
it's the variety he hears around him.  The same thing happens in 
English, as linguists studying non-standard English (typically Black 
English) have been at pains to point out.

There is of course language change, where one generation makes the
language slightly different from the previous language.  Often this
takes the form of leveling irregularity, or changing forms in one
paradigm class to another based on analogy [and of course sound change].

> I don't have a (strong) opinion on whether language facility is 
> built-in, but perhaps we should be taking about low barrier of entry 
> to language (any negotiated noise/signal would have an immediate 
> pay-off), as opposed to a high barrier of entry to more abstract 
> sciences where anything useful requires learning a large body of 
> knowledge first.

Language does not have a low barrier.  Just ask any second language
learner, or for that matter any linguist who's tried to write a grammar.

>> Contrast this with the learning of math or science or history,
>> where even with years of intensive teaching, not everyone learns
>> it.
> 
> Again, isn't it because most of the math/science/history is somewhat 
> abstract and does not engage the same centers of the brain?

Grammar is abstract too.  But I will agree that it doesn't use the same
brain centers--and that's the point: first language learning is
different, and uses specialized brain centers that appear to have little
to do with "ordinary" learning.

I didn't originally reply to Alexandre's spitting example, but because 
this has come up again: hitting a nearby target does not require 
physics, it just requires some practice (and sometimes a good wind). 
Maybe the nearest thing in language would be mimicking the sound of 
someone else's language.  Most people can't do it well (nor can most 
people win spitting contests), and those who do certainly can't be said 
to have learned the language.
-- 
    Mike Maxwell
    "We signify something too narrow when we say:
    Man is a grammatical animal. For although there
    is no animal except man with a knowledge of grammar,
    yet not every man has a knowledge of grammar."
    --Martianus Capella, "The Seven Liberal Arts"

_______________________________________________
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora