Yuri
Tarnopolsky
This paper continues the examination of language as a quasi-molecular system from the point of view of a chemist who happens to ask, “What if the words were atoms?” Any new word in the vocabulary must be at some time heard or read in order to be acquired. The new word practically always comes linked either with an observed image or with a few other words in a phrase or discourse, otherwise it would be meaningless at the first encounter. This halo of selective connections makes the morphemes and words recognizable as generators of Pattern Theory (Ulf Grenander), i.e., as atomic objects possessing a certain structure of potential bonds with preferences for binary coupling. In this way the word is typically acquired with a fragment of grammar. Metaphorically speaking, the generators of language carry bits of grammar on their bonds like the bees carry pollen on their feet. Regarding Pattern Theory as meta-theory for atoms and words, parallels between linguistics and chemistry are discussed.
CONTENTS
Introduction
1. Sets and Order
2. Laws of grammar and laws of nature
3. From sets to generators
4. Patterns
5. Notes on notation
6. Bonding
7. Acquisition of generators
8. Bond space
9. Acquisition of bond values
10. Locality
11. Some examples
12. Language and homeostasis
Conclusion
References
2004
Keywords:
linguistics, language, grammar, generative linguistics, poverty of
stimulus, language acquisition, language evolution, language and
chemistry, atoms and words, molecules and phrases, Pattern
Theory, Ulf Grenander, Noam Chomsky, Mark Baker, Brian MacWinney,
Manfred
Eigen, Ilya Prigogine, Walter Ross Ashby, Martin Novak.
© Yuri Tarnopolsky, 2004