Re: a request / Connectionism

Gordon Wells (gwells who-is-at oise.utoronto.ca)
Wed, 18 Mar 1998 09:58:43 -0500 (EST)

On Tue, 17 Mar 1998, John St. Julien wrote:
>
> I find a lot of this reasoning pretty convincing. But only if you start
> with the assumption that language is generated by rules. (It is a different
> thesis to say it can be _described_ by rules.)
>
> Connectionists and the larger group of those who also turn to a dynamic
> style of analysis would disagree with this premise and claim that the
> patterns we describe as rules are an emergent regularity of a
> self-organizing system. This is where the past tense debate came in. The
> two groups had different explanations of why children go through a period
> using past tense competently in their limited vocabulary but then seem to
> loose that competence as their vocabulary expands. Rules vs. experience was
> at the heart of the differing explanations.

Although at the time I didn't put it that way, the conclusion I reached
from my longitudinal study of a representative sample of 128 children was
rather similar. In 1969, following the lead of Roger Brown, I started
the study with Chomskyan expectations, but over the next fifteen years the
evidence persuaded me that implications of an innate LAD equipped with a
universal grammar were untenable. Certainly, there was strong similarity
between the children in the sequence of development, just as there was in
the relative frequency of the patterns learned in the speech addressed to
them. But just as striking were the differences between children and
these were also strongly related to the differences in experience between
them. At the time, I found Vygotsky's theory of learning and development
the most helpful in making sense of these results.

I have recently been reading two books that have helped me further. The
first is Yngve's "From grammar to science: New foundations for general
linguistics", which argues strongly against the Chomskyan notion of
speech being generated by rules. The second is Terence Deacon's "The
symbolic species: The co-evolution of language and the brain." He, too,
puts forward a strong argument against an innate language "organ" and
argues instead for an emergent co-organization of both brain and symbolic
activity. Although neither book has an explicit CHAT orientation, both
are in general compatible with it.

Deacon, T. (1997) The symbolic species: The co-evolution of language and
the brain.

Yngve, V. (1996) From grammar to science: New foundations for general
linguistics. Amsterdam/Philadelphia: John Benjamins

Wells, G. (1985) Language development in the pre-school years. Cambridge:
Cambridge University Press.

Wells, G. (1986) The meaning makers: Children learning language and using
language to learn. Portsmouth, NH: Heinemann.

Gordon Wells, gwells who-is-at oise.utoronto.ca
OISE/University of Toronto
http://www.oise.utoronto.ca/~ctd/DICEP/