Re: THEORY nouns and cases (was: Verbs derived from noun cases)

I know that some natlangs partially do so, in a rather
incomplete, ambiguous and illogical way (because they
are natlangs). But I understand Lojban is a conlang,
and more, a logical conlang.

Let me just quote a sentence of Oracle documentation I
came across this very morning:

"A data model for any application is comprised of
entities, associations among these entities, and
attributes that describe both the entities and the
associations".

Well, looks to me that this is very similar to a
language. We have entities (essentially, "nouns", at
least, "real" nouns, not deverbal ones, for ex); we
have associations (I would have said relations; those
are essentially "verbs", but of course we have to me
more precise about this term further), and we have
"attributes", or "properties" (adjectives may be
considered as noun properties for ex, and adverbs as
verb or whole-predicate properties).

Entities, relations and properties are either defined
in the lexicon, either in the speech. (There should be
some more concepts though, like modals for ex).

The brown fox jumped playfully over the lazy dog:
- "fox" is an entity
- "brown" is a (speech) property of "fox"
- "to jump" is a relation (in this case, a movement
action)
- playfully is a (speech) property of "to jump"
- (over the lazy dog) is a circumstant, which has to
be further decomposed
- "is-a-mammal", "is-furry", are implicit relations of
"fox", defined in the lexicon

So I of course could try to rewrite this sentence like
for ex: "A playful jump brownfoxed over the lazy dog",
which wouldn't be shocking syntactically, but it looks
very weird semantically.

This is of course very schematic, but it's the idea.

--- Tim May <butsuri@...> wrote:
> Philippe Caquant wrote at 2004-04-26 12:14:40
> (-0700)
>  > I agree with "to dog" or "desum", and many more
> (in
>  > French we haven't a verb "to dog" (chienner ?),
> but we
>  > have "singer" (to imitate mockingly, like a
> monkey). I
>  > said that nearly anything could become a verb,
> but my
>  > question was: zachem ? as the Russians say, "what
> for
>  > ?"
>
> What do you mean, "what for"?  There are natural
> languages which do
> the same thing, or very close (theorists disagree).
> See
> e.g. Nuuchahnulth (Nootka):
>
http://wings.buffalo.edu/soc-sci/linguistics/people/students/dissertations/davidson.PDF
>
> For example
> |
> |In Makah and Nuuchahnulth, nominals may function
> directly as predicate
> |heads with no intervening copular element.  They
> take predicate
> |clitics exactly as verbal predicates do: there are
> no restrictions on
> |the predicate clitics they may occur with.  The
> words in (205)-(206)
> |show Makah nouns and property words as heads of
> class-inclusion
> |predicates;  that is, predicates denoting a class
> of entities the
> |subject is asserted to be a member of.  Example
> (207) shows an
> |intransitive verbal predicate for comparison.  The
> coding of each
> |clause type is identical: the mood and pronominal
> clitics are attached
> |directly to the predicate head in all cases (only
> the masculine
> |singular gloss is given for the third person
> examples for sake of
> |economy.)
> |
> |          MAKAH
> |(205) a.  wikwi'ya'ks	    b.  wikwi'ya'wic	c.
> wikwi'ya'w
> |	   wikwi'ya:k^w=s       wikwi'ya:k^w=°ic
> wikwi'ya:k^w=°i
> |	   boy=INDIC.1sg        boy=INDIC.2sg
> boy=INDIC.3sg
> |	   'I am a boy'         'You are a boy'     'He is
> a boy'
> |
> |(206) a.  k^wa?aks	    b.  k^wa?awic	c.  k^wa?aw	
> |	   k^wa?ak^w=s	        k^wa?ak^w=°ic	
> k^wa?ak^w=°i
> |	   small=INDIC.1sg	small=INDIC.2sg
> small=INDIC.3sg
> |	   'I am small'         'You are small'     'He is
> small'
> |
> |(207) a.  babuyaks	    b.	babuyawic	c.  babuyaw
> |	   babuyak^w=s		babuyak^w=°ic	    babuyak^w=°i
> |	   work=INDIC.1sg	work=INDIC.2sg
> work=INDIC.3sg
> |	   'I am working'       'You are working'   'He is
> working'
>
>  > In order to hide under the carpet the fact there
> are such concepts
>  > as entities, properties, relations, temporary
> states, etc ? Why
>  > should we pretend that entities (choses-en-soi)
> are similar to
>  > action verbs, for ex ?
>
> Why should we pretend they aren't?  What makes you
> think their
> differences are so much more significant than their
> similarities that
> they should be placed in lexicalised syntactic
> classes?  There's no
> unarguably correct way of dividing up concepts into
> noun/verb/adjective; natural languages differ as to
> which category
> things should be assigned, even when the languages
> have similar sorts
> of category.

=====
Philippe Caquant

"High thoughts must have high language." (Aristophanes, Frogs)


	
		
__________________________________
Do you Yahoo!?
Win a $20,000 Career Makeover at Yahoo! HotJobs
http://hotjobs.sweepstakes.yahoo.com/careermakeover

Reply