Re: THEORY: Parts of speech, was: Syntaxy-Turvy

>===== Original Message From Constructed Languages List
<CONLANG@...> =====
>>I see the (conceptualized) world as consisting of an unspeakably complex
>>webwork of entites linked to each other; some of these entities are
>things,
>>and others are states/processes/events,
>
>Therefore, we agree here: the reality (or the cells of our conceptual
>grids) CAN be classified into dynamic processes, static situations,
>more or less stable objects, qualities, etc.
>
>But then, lexemes in *any* language can be classified this way. The
>difference is rather if/how this classification is used in a specific
>language to sort words into groups worth terming them 'parts of speech'.
>
>If you proceed from lexemes like I did above, it is only natural to
>call 'nouns' or 'substantives' the class where most terms denoting
>stable objects belong, whatever their syntactic/paradigmatic qualities.
>
>It seems that syntactic classification of words is much less uniform
>across languages, and can be unrelated to the 'lexemic' one.
I doubt it is ever completely unrelated, but there certainly can be
considerable mismatch.  I doubt you'll ever find a natural language which
doesn't, on the whole, have sentences which are structured around a central
verb, explicit or implied, and its one or more attendant nouns, explicit or
implied.  (In this sentence I mean 'verb' and 'noun' in their semantically
defined senses.)

But then again, I'm in very unfamiliar territory with this whole project.  I
usually think of linguistics in terms of people like Ronald Langacker, who
defines *everything* semantically, and does not believe in autonomous syntax
at *all*.  To accomplish this he has to use some extremely subtle (a detractor
might say ad hoc and bogus) semantic factors, and he also has to work not just
from the semantics of *words* but from the semantics of *constructions* -- for
example, he would recognize a "sentence" as a construction (with conceptual
status no different from any other construction, such as the noun phrase, the
passive sentence, or the interjection).  To Langacker, "syntax" is purely
epiphenomenal, a result of the interaction of different constructions with
each other.

I came up with Taxy after reading a book that, while radical in some ways,
does to some degree recognize "syntax" as a thing unto itself.  But I'm not
used to thinking about syntax independently, so I'm a little unsure of myself.

I'd best keep working on it and see how it works itself out in practice.

>I suspect very much that this is merely a function of stability.
>It seems that most features attributed to 'lexemic substantives' are
>actually included in the concept of the class of objects denoted by
>the given word (e. g. _people_ typically have _heads_, _legs_, etc.).
>More dynamic entities (denoted by 'lexemic verbs') have more 'modular'
>structure, that is, their important features are participants different
>in each particular instance. In other words, an object's instability in
>the real world correlates with the number of details that must be
>attributed to it, to obtain a meaningful description, on the level of
>a sentence (a dynamically generated linguistic entity) rather than
>simply definition of the term (a relatively fixed linguistic entity).
I'm not sure I agree with you here.  You're saying that the reason "eating" is
a verb and "sandwich" is a noun is that "eating" in the real world is more
unstable than "sandwich" in the real world?

>This looks correct.
>
>But: it seems that you define parts of speech as syntactic classes,
>which leads to a question more difficult for me: what are 'nouns',
>'verbs', etc. from a purely syntactic point of view?
I don't think that in Taxy I define parts of speech as syntactic classes.  I
define them semantically -- nouns are prototypically objects, and verbs are
prototypically activities.

The difference between English and Taxy is that where English takes verbs
(semantically defined) as the core of a sentence, and surrounds them with
whatever nouns (semantically defined) they are associated with, Taxy takes
nouns (semantically defined) as the core of the sentence, and surrounds them
with whatever verbs (semantically defined) they are associated with.

>I feel that you invented a totally original system, but I'm trying to
>understand what exactly is 'inverted' in it, as compared to e. g.
>English.
>
>(And I think you'll still need more diverse markers for what is
>attributed to 'lexemic verbs' compared to 'lexemic nouns'.)
Well, I'll work a bit more on it, incorporating Christophe's excellent
suggestions, and post it again; perhaps it will be more clear.

-------------------------------------------
         edheil@mailandnews.com
-------------------------------------------