Re: ergative? I don't know...

At 7:45 PM -0500 10/24/98, Sally Caves wrote:
>On Fri, 23 Oct 1998, David G. Durand wrote:
>> A (most agent like argument, subject of an transitive verb), P (most
>> patient-like argument, Object of an intransitive verb), S (Subject of an
>> intransitive verb). S may be close to your "Experiencer", though that term
>> is also used for something related but diffrent.
>>
>> These are often marked somehow by morpholoy or syntax, into cases:
>>
>> i. nom/acc system:
>>
>> Nominative  (A, S)
>> Accusative  (P)
>
>This is how I think I am classifying Teonaht:  it has a nominative that
>can be further subdivided into two categories:  Volitional nominative and
>non-volitional nominative.  So, as I have ad nauseam explained, you say
>
>                Il jentwar le zef elo cosa
>                "the door (P) the man (A) closed,"
>
>or: Le zef il jentwar elo cosa  "the man the door closed"
>
>And "close" would be a volitional transitive.
>
>You can also say:
>
>                Le uehar oua,
>                "the woman (A) listened,"
>
>        and "listened" would be a volitional intransitive.
This sounds like the way I'd be tempted to classify it, especially as the
nominative seems to handle agents (A) in transitive clauses, and also S
(subject of intransitive). Splitting S arguments but agent-worthiness is
certainly done. I've read of it in some "active languages" (ones with 3
cases for A, P, and S). Of course if you split this function in an active
language, you end up with _four_ cases, unless you re-use your "A" case on
intransitive verbs as well. You definitely don't do that, since your P case
is never used as a subject.

What seems interesting to me, is the passivization process where you use an
S-marked (unagentive) subject of a transitive clause, like this:


>                But:
>                Li zef il jentwar nelo ke lis ai aid cosarem
>                "the man (S) the door (P) he saw (P) it get its closing"
>                because the man is in a passive state of seeing.
>                "the man saw the door get closed, or close."
I notice that the subject is echoed as a pronoun (he). What case is that
echo in?

>As yet, I don't have a real middle voice, although I think one is
>beginning to develop from this passive:
>
>        The door closes.  This would be expressed in T. with a "passive":
>
>                        Li jentwar tsobccosarem [na]
>                        The (S) door under closing [is]
>
>                The copula is left off in the present tense,
>                and the tendency in this idiom is to omit as well
>                the gerundive suffix:  Li jentwar tsobccosa
>                                        "the door underclosing."
>
>                This is reconceived by T. speakers as a kind of
>                "middle voice verb," so that any verb with prefix
>                "hsob/hsop" is de-passified:  "the door undercloses."
>                New verb: tsobccosaned, "to be under closing," i.e.
>                "to be closed," "to close."
This is why I didn't go into voice phenomena. I've now read several hundred
pages on argument marking and voice, since I got interested in it a year
ago. Even Payne's very clear (but occasionally simplified) discussion of
argument marking is about 39 pages. He's got another whole chapter (53
pages) on voice, causatives, and other ways of changing the argument
structures of verbs. Your Teonaht system seems naturalistic to me, whether
or not it's attested anywhere else. Matt may have an idea of other
languages that might use such a system. If Mark Line were still talking
here, he might well have useful data or opinions. (If you're lurking, Hi
Mark!).

>But here I draw away from the point, which is that there are two types of
>NOMINATIVE in this nom/acc. language which make a distinction based on
>volitionality.  Also, the P can NEVER function in the subject slot.  I.e.,
>Teonaht does NOT say il jentwar tsobccosa(rem), "the door (P)
>underclosing."  Above, even though the door is actually getting closed,
>like a true nom/acc language, it is cast in some form of the nominative
>(non-volitional).
>
>Again I ask...What other natlangs do this?  [dead silence]
It seems sensible, (certainly easier to grasp than the "inverse systems"
that are based on the animacy hierarchy).


>> iii. active system
>> Agent       (A)
>> Patient     (P)
>> Subject     (S)
>
>Do these constitute three separate cases?  I don't know if T. fits this.
These would be 3 separate cases. I was trying to list cases on the left,
and roles on the right, with case names chosen as traditional within that
kind of system. I guessed on the names of the cases for active systems
'cause I couldn't remember

>> Some active (and ergative languages) may differentiate "agent-worthiness"
>> in the S argument, so that intransitive sentences like "John kills" and
>> "the door closes" use different cases to reflect the difference in
>> agenthood.
>
>Definitely:  In T. "kill" would be "volitional transitive" and John would
>be an agent or a volitional subject; but "closes" would be nonvolitional
>intransitive, or it would be in the passive, and door would be a
>participant, or non-volitional subject.
Right. It sounds to me like this is close to what Teonaht is doing. I'd
guess that many instransitive sentences happen to be less agentive, and
many trnasitive sentences more so (the unagentive nominative has got to be
rare with "kill" except for accidents or serial killers).

>> Some languages use one system (ergative or accusative) in one tense, and
>> another in another tense (e.g. past tenses erg/abs, non-past nom/acc), some
>> use one system on bound verbal pronouns, and the other on fully expressed
>> nouns. These are the "split" systems.
>
>These combinations sound just as strange and unnatural as the gyrations
>I'm putting Teonaht through.
To us... That's the wondrous thing about language. Your system sounds
speakable to me, so I wouldn't worry about whether it's attested or not. I
hope you get your web pages finished...

>> 1p > 2p > 3p > animate (present) > animate (absent) > inimate (present) >
>>     > inanimate (absent)
>
>inimate?  intimate or animate?
inanimate. Intimate is more likely to show up on a formality hierarchy...
>
>This is a rare type, but is found in some Algonkian languages (is
>> that the proper modern spelling?).
>
>Algonquin, I believe.
I think that's a traditional spelling for the cultural group. I know
Bloomfield and the structuralists talked about Algonquian languages. And
recently I've seen  titles that refer to Algonkian -- I was wondering if
they've changed to orthography... They just changed the orthography for
Mayan (tribal decisions to make it more natural and less clearly
spanish-derived).

Could just have been a weird linguist though.

   -- David

A random series of thoughts triggered by your old surveys prompted me to
spew the following lines into the PS part of my email.

One change that CONLANG made in my life is motivating me to finally really
work through some of the parts of syntax that I found lastingly confusing.
Of course, I taught myself most of my linguistics from old structuralist
texbooks and grammars of North American Indian languages in high school.
Syntax was never  their strong point, though, and when I reached college, I
took a few linguistics courses. In Syntax I they taught us from Chomsky '67
and a general TG textbook, as if the theory was even worth the value of the
paper it was printed on. As a computer geek, I could not believe in a
syntactic formalism that was:
  a) provably unparseable (context sensitiveby the second week, turing
complete by the thrid week)
  b) still couldn't handle everything clearly once it had reached the power
of a universal turing machine, but still needed patches on the patches for
the analyses to look remotely plausible.

Of course they started the following semester (Syntax II) by saying "Now
you're going to learn why everything we did last semester is wrong." And
they then introduced some of the competing theories like GPSG, Unification
Grammar, and other stuff. Too bad for me that I'd given up on linguistics
already (if that's what linguists did, I wanted no part of it). I didn't
learn that there even _were_ other serious theories until I was about to
graduate, and talked to someone who didn't drop syntax after that first
semester. I think I make a better CS guy than I would have a linguist, but
it was an irritating experience.

And I never learned any syntax that I could believe except for the
(moderately hackish, and very English-centric) stuff you do in traditional
AI classes. I know English grammar pretty well, and so it didn't really
help me to learn general syntactic principles the way reading structuralist
grammars had given me a good sense for morphological principles.

CONLANG (and Matt's recommendation of the 3-volume collection by Timothy
Shopen) got me back into syntax by means of typological and descriptive
linguistics. I was wandering in this section of the Brown library about a
year ago (and that was where, a few stacks over, I had used to read
linguistics in high school), seeing a whole new world of grammatical
information about languages I'd never heard of, with explanations of how
their syntaxes actually functioned!

I think it's time I let thesis avoidance go again, and get back to
change-oriented concurrency control and collaborative editing.



_________________________________________
David Durand              dgd@cs.bu.edu  \  david@dynamicDiagrams.com
Boston University Computer Science        \  Sr. Analyst
http://www.cs.bu.edu/students/grads/dgd/   \  Dynamic Diagrams
--------------------------------------------\  http://www.dynamicDiagrams.com/
MAPA: mapping for the WWW                    \__________________________