Chomsky

[Keith Johnson]

Hi Sheri,

I wonder if your feeling that your daughter had crossed a language  
learning threshold might have been at about 18 months, the usual time  
of the "vocabulary explosion"?

Bob McMurray has some really interesting ideas about how to explain  
this feature of language acquisition. He simulated the vocabulary  
explosion by making only two assumptions.  (1) words are being learned  
in parallel - a little bit of learning for many words at the same  
time, and (2) some words are easier to learn than others.  That's all  
it takes to have a vocabulary explosion - no language module needed.

McMurray posted a very helpful discussion of his work:
http://www.psychology.uiowa.edu/faculty/mcmurray/science/

If the sense that a child has crossed a linguistic threshold is  
related to the vocabulary explosion, and if the vocabulary explosion  
is related to McMurray's two factors, then what keeps Alex from  
crossing the threshold must be one of two things; he is learning words  
sequentially rather than in parallel, or he finds it much harder to  
learn words than children do.  If the first impediment is the culprit  
then perhaps training methods could be adjusted, but if the second is  
the crucial factor then it may not be possible for a parrot to cross  
that language-learning threshold with English.

Earlier work with Chimps has tried to address the word difficulty  
problem by teaching ASL rather than spoken language, but one wonders  
if the difficulty lies at a more conceptual level of word learning  
than mere input/output system unnaturalness.  Still the LACUS paper  
that Aya pointed us to hints at an impressive use by parrots of  
duality of patterning (one of Hockett's 13 design features of language  
Hockett, C (1960) "The Origin of Speech". in Scientific American, 203,  
pp. 89–97).

best,
Keith Johnson

p.s. I would like to say, regarding the range of human linguistic  
ability (Aya's repeated point that some humans don't have language),  
that it would be useful to remember that clinical speech and language  
pathologists have developed numerous standardized tests of human  
linguistic ability, and school-aged children are routinely tested to  
detect developmental delays.  Perhaps, if one wanted to measure the  
level of linguistic accomplishment reached by a parrot or other animal  
it would be good to score the animal on some standardized tests.