Re: Fw: Intro to natural language processing
John,
I agree with you about the importance of relating
NL semantics to psychological and neurophysiological
evidence:
JB> For a nice, understated and empirically well grounded
> discussion and construction of models involving
> spatial language and embodiment, see:
>
> Saying, Seeing and Acting: The Psychological Semantics
> of Spatial Prepositions (Essays in Cognitive Psychology)
> by Kenny R. Coventry, Simon Garrod (January 1, 2004)
An abstract of a talk by Coventry is available on the
following web site (along with abstracts of talks by
other people on related issues):
http://www.kyb.tuebingen.mpg.de/misc/friko/
This kind of work is very compatible with Len Talmy's
cross-linguistic studies of spatial relations, and
a multidisciplinary convergence seems quite likely.
I'd like to mention another book that combines neuro-
physiological evidence with anthropological studies
and Peirce's semiotics:
Deacon, Terrence W. (1997) The Symbolic Species:
The Co-evolution of Language and the Brain,
W. W. Norton, New York.
I am in the process of writing a review article on some
related topics, and I include a brief summary (copy below)
of Deacon's position. I believe that Deacon's arguments
are much more cogent than Jackendoff's (who criticizes
D. along Chomskyan lines). In the interest of giving
equal time, following is the URL for Jackendoff's own
summary of his book:
http://www.bbsonline.org/Preprints/Jackendoff-07252002/Referees/
Although Jackendoff diverges from Chomsky, he still
believes in some form of Universal Grammar (UG) in the
genes. Deacon makes that view seem very doubtful.
John Sowa
___________________________________________________________
From a forthcoming article by J. F. Sowa:
Chomsky and his followers, such as Jackendoff (2002), maintain that
children learn language so rapidly because the human brain has an innate
module that facilitates the acquisition of syntax. After studying human
evolution, Deacon (1997) agreed that the brain has a predisposition to
language, but that the evidence shows a slow evolution of the brain and
vocal tract toward modern forms. It suggests that early hominids already
had a rudimentary language, which gave individuals with larger brains
and better speaking ability a competitive advantage. To analyze human
and animal communication, Deacon used Peirce's semiotic categories of
icon, index, and symbol to classify the kinds of signs they could
recognize and produce. He found that higher mammals easily recognize the
first two kinds, icons and indexes, but only after lengthy training
could a few talented chimpanzees learn to recognize symbolic
expressions. Deacon concluded that if chimpanzees could make the
semiotic transition from indexes to symbols, early hominids could. Once
that transition had been made, language was possible, and the use of
language would have promoted the co-evolution of both language and brain.
Although Jackendoff diverged from Chomsky by assuming graph-like
conceptual structures, he replied to Deacon by repeating the claims for
innate syntax. But if language depended on an innate syntactic module, a
simulation of that module might enable computers to learn and understand
language as easily as children. Yet syntax is not what makes language
difficult for computers, and it's not essential for human understanding.
People speaking a foreign language can make themselves understood even
with badly fractured syntax, which is notoriously difficult for
computers to parse. That supports Deacon's hypothesis that symbols are
more important than syntax: at first, children learn words as symbols
for things; their predisposition to mimic adult speech enables them to
learn patterns of words; and syntax results from learning to map speech
patterns to cognitive patterns. Contrary to Chomsky, universal grammar
is not the prerequisite, but the result of many generations of children
preferring patterns that are easy to learn. Further research in
semiotics may clarify these issues; promising techniques include
algebraic semiotics (Goguen 1999), a dynamic logic of semiosis (Deacon
2004), and a combination of Peirce's semiotics with conceptual
structures (Sowa 2000).
The .PPT slides that Deacon used for his 2004 talk are available
in a 10 megabyte download:
http://www.cs.uah.edu/~delugach/CG/ICCS04/ICCS04-Web/DeaconHuntsville04.ppt