Reflections cont'd (2)

[Salinas17 at aol.com]

I wrote:
<<In response, I only can tenderly carmelize the teacher to whom you 
ventilated in the left column, weedlessly uncondensed.  That's fine grammar but -- 
unless you are privileged to know something I don't -- it makes absolutely no 
sense.>>


In a message dated 3/17/06 7:56:12 PM, mark at polymathix.com writes:

<<While that is presumably true in the context of a linguistics mailing list, 
I doubt there's much evidence of the same kind of phenomenon in the 
populations of mere mortals and their utterances that we purport to study. >>

No doubt that it is an improbable sentence.  And it just happens to make no 
sense as well.  Maybe they are connected?  

This goes back to your suggestion about language making reality intelligible. 
 My point is it's more useful to see it the other way around.  That reality 
makes language intelligible.  And when reality doesn't check-off on a sentence 
like the one above, it will tend not to make sense and therefore tend not to 
be used.

If you turn Chomsky on his head -- "Colourless green ideas sleep furiously" 
-- you may think what he proved was that language that makes no sense is not 
language.

What does that suggest about "language models?"  It might suggests that any 
model of language that does not include the extra-linguistic effects of 
language is fundamentally inaccurate.

Regards,
Steve Long