[Corpora-List] Chinese and English POS

[Andras Kornai]

On Tue, Nov 03, 2009 at 11:24:09PM -0500, Mike Maxwell wrote:
> In addition, an analysis of a particular language might give you ways to 
> figure out what PoS a particular word in a particular context is.  

I think part of the problem is the universality, or lack thereof, of
the part of speech categories. For those of us accustomed to
IndoEuropean languages in general and English in particular, the
universality of at least the major POS categories, nouns, verbs,
adjectives, and adverbials, *appears* as inevitable. Chinese really
desn't fit the IE scheme well in that what we call (from the IE
perspective) verbs and adjectives are essentially indistinguishable
from one another. The right POS distinction seems to be between these
"void words" on the one hand and "solid words" (nouns) on the other.
Void words are linked as predicate to nouns (solid words) by
juxtaposition, and are negated by the same particle pu (note that in 
English verbs and adjectives are negated differently, "the man doesn't 
work" "the boy isn't tall") while nouns are linked to nouns by means of 
a copula yeh or shih, and are negated by fei or pu shih at least according
to AC Graham's "Two Chinese Philosophers" (p41 of the Open Court edition). 

Graham also provides wonderful examples of Chinese words with a
clearly unitary meaning that can function both as transitive verbs,
intransitive verbs, and adjectives, e.g. sheng "to give birth to
(child)" "to grow (hair, boils)" are transitive, "be born" or "be
grown" are clearly intransitive, and "raw" (for food, as opposed to
cooked) is clearly adjectival (p 47). I know very little about
Chinese, but in my native Hungarian these would be derived from the
same stem or root but distinguished from one another by means of
category-changing suffixes.  For example English "divorce" (noun)
would be val-as while "divorce" (verb) would be val-ik [I apologize
for the lack of diacritics throughout]. 

Whether a category distinction is present, just not signalled by the
language (as with English "divorce") or not even present (as with
Chinese "sheng") is generally not a question that can be settled in
isolation. We grit our teeth and acknowledge "divorce" as a noun-verb
precisely because the distinction between nouns and verbs is supported
by some many forms of evidence in English, with zillions of cases that
are only nouns or only verbs. However, when the structural evidence
for distinguishing adjectives and verbs is missing, as in Chinese,
there doesn't seem to be any point in pretending every stem in the
void class is ambiguous. 

As for the larger debate between Mike and Geoff, there is no ending
this, see Robbins Burling: Cognition and Componential Analysis --
God's Truth or Hocus-Pocus. American Anthropologist, 66 (1964), 20-28

Andras Kornai

_______________________________________________
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora