"H. S. Teoh" wrote:

> On Wed, Sep 27, 2000 at 01:52:55AM +0200, Jörg Rhiemeier wrote:
> > "H. S. Teoh" wrote:
> [snip delicious comparison of cases in my conlang with Nur Ellen]
> >
> > This is indeed an interesting case system!  Very original, and a
> > relational typologist's nightmare ;-)
> I guess the fact that I like thinking "out of the box" may have something
> to do with this :-)
> I notice that your case system carefully distinguishes between animacy and
> non-animacy.

This it does.  Inanimate things of course have no volition, and hence
there is no agentive, and some preopositions cannot be used with them
either.  These prepositions are the possessive genitive, dative,
comitative, benefactive and malefactive prepositions which all require
the noun they govern to be in the agentive case, because inanimate nouns
make no semantic sense.

>       The POV of my conlang is more leaned towards the semantic
> function of the noun in the context of the event, regardless of animacy or
> volition on the part of the noun itself. So, a fountain gushing up water
> would be marked in the originative the same way as a man speaking words,
> or a place out of which something else moves.

Yes, the views reflected by the case systems of our conlangs are quite
different.  Of your three originative examples, the man would be marked
agentive, the other two ablative.

And yet both reflect deeply-rooted semantic relations rather than
merely formal "subject/object" functions!

> [snip]
> > > BUT:
> > > 2)      mir33'n0      fww't3    pii'z3du.
> > >         children(org) see(verb) man(rcp)
> > >         "The man sees the children."
> > >         The object being seen is regarded as the transmitter of its own
> > >         appearance; hence the strange reversal of the expected cases.
> >
> > Nur-ellen treats this differently.  This is a typical sentence
> > of perception, where the perceiving entity goes into the dative
> > and the object of perception into the objective:
> >
> > Na  i   ben     tir i   jin.
> > DAT the see the OBJ.child.PL
> Very interesting! It does seem to resemble the way my conlang handles the
> verb to see, at least as far as using the receptive (dative) case is
> concerned.

Yes, there is a parallel in it.  Here, our langs are actually pretty
close to each other.  The main difference is that you cast the object
of perception in an agent-like originative role, while in my lang,
it is not actively involved and thus takes the objective case.

> > Without the dative preposition, the sentence expresses deliberate
> > observation:
> >
> > I   ben     tir   i   jin.
> > The watch the OBJ.child.PL
> > "The man watches the children."
> My conlang uses another verb, "zota'", "to look at", to express this. It
> might interest you to know that this verb uses rather different case
> markings from "fww't3":
>         pii'z3d0 zotuw'     mir33'nu.
>         man(org) look(verb) children(rcp)

The cases are flipped in comparison to _zota'_.

>         "The man looks at the children", or, "the man looks in the
>         direction of the children" (hence children in the receptive case:
>         they are the destination of the man's looking.)
> (Just in case you're wondering, "zotuw'" is the perfective deliberative of
> "zota'". I didn't indicate this in the interlinear to avoid clutter and
> possible confusion. If you *really* want to know, though, the deliberative
> must be used here to properly convey the meaning of "watch" -- if I used,
> say, the incidental, the sentence would read "The man happened to look at
> the children" -- i.e., he isn't actively watching them, but just
> happened to look in their direction.)

It is this deliberative/incidental distinction that is expressed by
choice of agentive vs. dative in Nur-ellen.  (The dative also has
more straightforward uses; see my translation of your example 1) and
the comment below.)

> [snip]
> > > Another one of my favorite pathological (from the POV of subject/object
> > > systems) cases:
> > > 5)      byy'jh     pii'z3da   3lymo3'n     biz3tau'.
> > >         give(verb) man(instr) flowers(cvy) woman(rcp)
> > >         "The man delivers the flowers to the woman."
> > >         Why is the man in the instrumental case? Because he is the person
> > >         delivering the flowers sent by somebody else. If "man" is put into
> > >         the originative, the sentence becomes "The man gives flowers to
> > >         the woman".
> >
> > In Nur-ellen:   instrumental (_ni_ + objective)
> >               + objective
> >               + benefactive (_an_ + agentive).
> >
> > Ni   i   ven     an`n i   ljös          an  i   bes.
> > INST the give the OBJ.flower.PL BEN the AGT.woman

Instead of the benefactive one could also use the dative here.
This is the "more straightforward" use of the dative I mentioned
above.  The difference between dative and benefactive is that
the benefactive implies that the person marked this way actually
benefits from the action, while the dative is neutral to this
regard.  There is also a malefactive, which expresses that the
person suffers from the action:

I   ben     hed   bomb dan i   bes.
The throw bomb MAL the AGT.woman
"The man throws a bomb at the woman."

The dative would be equally possible here.

> > Nur-ellen agrees here with your language by using the instrumental
> > for someone acting on behalf of someone else.
> Very interesting! My conlang looks at it from another angle, though: the
> instrumental is used for nouns that are responsible for carrying through
> with the event. In this example, the person delivering the flowers is
> important to the completion of the event of giving -- without this person,
> the flowers would not have been delivered to the woman.

This is generally the same the instrumental is used for in Nur-ellen:
an entity, animate or inanimate, by which something is carried out,
but that is not identical with the entity who causes this something
to be carried out.

> Similarly, the instrumental is also used for the vehicle of motion, when
> used with verbs of motion, because it is the vehicle that keeps the motion
> going.

Yes, the instrumental is also used for this in Nur-ellen.

> > Let's summarize the correspondences observed:
> [snip summary of comparisons]
> > It seems that no regular match can be observed.  The only correspondences
> > occuring twice in these samples are rec./dat. in 1) and 2),
> > and conv./agt. in 3) and 4).  The correspondence of the instrumentals
> > with each other in 5) is striking.
> Yep. I'll be very interested to see how much the instrumental cases in our
> conlangs overlap :-)

In Nur-ellen, the instrumental is used for the following:

1. An instrument by which an action is performed:

   Fingolfin     dagnent   Mjorgoth    ni   meg`l.
   AGT.Fingolfin kill-PAST OBJ.Morgoth INST sword
   "Fingolfin killed Morgoth with a sword."
   (I know the fight went the other way in _The Silmarillion_,
    but this example just sprang to my mind.)

   This usage includes vehicles:

   Martin     rand    na Amerika ni   vilkir.
   AGT.Martin travel  to America INST airship
   "Martin travels to America by airship."

2. An inanimate agent, which in Nur-ellen is seen as an instrument
   by which someone else does something (not necessarily volitionally):

   Ni   i   gendel   gend    men   e        Garnil.
   INST the computer compute orbit GEN.PART Mars
   "The computer computes the orbit of Mars."


   I   men`lgoldir    gend    men   e        Garnil ni   i   gendel.
   The AGT.astronomer compute orbit GEN.PART Mars   INST the computer
   "The astronomer computes the orbit of Mars with the computer."

3. An animate entity acting on behalf of someone else:

   Ni   i   ven     an`n i   ljös          an  i   bes.
   INST the give the OBJ.flower.PL BEN the AGT.woman
   "The man delivers the flowes to the woman."

   The idea behind this usage is that the agent performs the role
   of an instrument by which someone else achieves something.

4. An animate entity acting under force, against its will:

   Ni   in  Ilj`l      dregent   o    i   Rjömin.
   INST the OBJ.Elf.PL flee-PAST from the OBJ.Roman.PL
   "The Elves fled the Romans."

   The Elves are in instrumental because they did not flee because
   they wanted to, but rather because they found themselves forced
   to flee by the Romans.

This is what I can recall to have thought of so far.

(The nouns in the examples above where no case is specified in the
interlinears are inanimate nouns which are always in the objective
case form.)

> [snip]
> > > So my question is... what *is* this system?? I've read a bit about active
> > > langs, ergative langs, and trigger langs, and my conlang doesn't seem to
> > > fit in any of those categories (although it seems closest to active).
> >
> > It seems to stand on its own.  It is none of the ones mentioned.
> > No trigger, but also not accusative, not ergative, and also not really
> > an active language.  At least, nothing I'd recognize as active from
> > what little experience I have with relational typology.
> > But interesting out of its own right.
> >
> > It makes the active system in Nur-ellen (or do you find a reason why
> > it is not active, Marcus?) look tame, even with its degrees of volition.
> Hahaha! Is my conlang really *that* off-the-wall? :-P

No, it isn't.  But it is very interesting, and unlike anything I have
yet seen and understood.

> [snip]
> > Nur-ellen also has sentences which consist of two nouns, or a noun and
> > an adjective, and the predicative nouns and adjectives are inflected
> > for tense!  (For an example, see below.)
> Cool! Well, my conlang can have stative sentences with more than two
> nouns, too, though I haven't worked out all the details yet.

You mean, predicative nouns can be transitive?

> [snip]
> > > Here's an example of a stative sentence: (cover story :-P)
> > >         biz3t30'   d3m3'l.
> > >         woman(org) beauty(cvy)
> > >         "The woman is beautiful." Literally, "the woman shows forth beauty".
> >
> > Nur-ellen:
> >
> > I   ves       vin.
> > the OBJ.woman OBJ.beautiful
> >
> > Nur-ellen is a zero-copula language; and as "the woman" isn't actively
> > doing anything in this sentence, the case marking is objective.
> Cool. I like that idea. I observe again that here, the case marking
> reflects the volition, or "activity" (activeness?) of the noun.

Yes.  No action -> objective.  Volitional action -> agentive.
Non-volitional action due to coincidence or error -> dative.
Cursory perception -> dative.  Non-volitional action due to an
external source of action -> instrumental.

> > As I said, predicative nouns and adjectives are marked for tense:
> >
> > Voromir     gondirent      e        Davrob`l.
> > OBJ.Boromir OBJ.mayor-PAST GEN.PART OBJ.Tavrob`l
> > "Boromir was mayor of Tavrob`l."
> >
> > (Yes, case and tense markers on the same word!  So here is the
> > 64,000 dollar question: is _gondirent_ a noun or a verb?)
> I'd say, it's something like a participle. After all, participles do have
> both case and tense markings (esp. in natlangs like Attic Greek) :-P

It's rather the inverse of a participle in being a noun taking on
verb inflection.  In fact, it is a noun with an enclitic zero
copula which is inflected for tense ;-)

> > This, however, implies that the condition does not hold in the present.
> > So if one talks about a women one has seen once, one would rather say
> > _I ves vin_ "The woman is beautiful" rather than _I ves vinent_
> > "The woman was beautiful"; the latter would imply that she is no longer
> > beautiful (because she has gained 50 pounds in the meantime, suffered
> > a terrible accident, had herself peirced, or whatever).
> My conlang would simply insert a temporal noun in the locative case, to
> indicate that it's talking about a previous state which may or may not
> hold in the present.

Such as "The woman [is] beautiful last Friday"?  In Nur-ellen, this

I   ves       vin           ad  Orbelt     vjedy.
the OBJ.woman OBJ.beautiful LOC OBJ.Friday OBJ.last

The present tense means that she might still be beautiful, while the
past tense would mean that she is no longer.

> [snip]
> > > Not all attributes are expressed using the originative-conveyant
> > > combination, however. Here's another attributive stative sentence:
> [...]
> > > This construct is known as the possessive stative. The conculture regards
> > > the thing possessed as being focused towards its possessor; hence, the
> > > possessee is in the conveyant case, and the possessor in the receptive.
> > > (Another way to think of this is that the possessor is the recipient of
> > > the possessee during some past acquisition.)
> >
> > An interesting view.  Just the reverse "direction" of "conveyance"!
> Not really. The conveyant noun is implicitly "moving" from the originative
> (source) to the receptive (destination). So when used with an originative
> noun, it marks a motion away from something; when used with a receptive
> noun, it marks a motion towards something.

In Nur-ellen, the ablative and adlative prepositions correspond to
your originative and receptive cases, while the conveyant itself takes
different cases or prepositions according to its degree of volition.

> There are more too... I'll post the details of that in the next tidbit :-)
> > It is always surprising which distinctions are made in languages.
> [snip]
> I find it very fascinating that Nur Ellen pays a lot of attention to the
> agency of its nouns. (Is this a common thing in active langs?)

I have the impression that it is, at least that is the way I understand
the concept of active languages.  Surely Daniel and Marcus can say
more about this.

>       The case
> markings depend very much on whether the noun is actively doing something
> (agentive), or is passive (objective), or neither (ablative, etc.). My
> conlang makes distinctions along different axes, so to speak.

Very interesting axes, indeed.