Re: Abandoning the Metaphysics of Subjects and Objects?
|From:||Sylvia Sotomayor <kelen@...>|
|Date:||Monday, August 25, 2003, 13:45|
On Sunday 24 August 2003 01:41 am, David Peterson wrote:
> So, looking at word order typology, we have the following:
> SOV = 45% of the world's languages
> SVO = 42%
> VSO = 9%
> VOS = 3%
> OVS = 0.9%
> OSV = 0.1%
> Comrie has pointed out the following:
> (1) Tendency for S to come before O: SO = 96% (SOV, SVO, VSO); OS = 4%(VOS, OVS, OSV)
> (2) Tendency for O and V to touch: Touching = 91% (SOV, SVO, VOS,OVS); not = 9% (VSO, OVS)
> I'd like to add the following to these:
> (3) Tendency for S and O to come together: SO = 57.1% (SOV, VSO, OSV,VOS); not = 42.9% (SVO, OVS).
> (4) Removing the extremely rare ones, you still get: SO = 54% (SOV,VSO); not = 42% (SVO).
> Now, to add a wrinkle: If one thought of subjects and objects as*identical* (i.e., the idea of agency is rendered unimporant), then you
could reduce the number of word orders to the following: NV, VN, NVN
(where N stands for "subject or object"). This would then be the make
up of the three dominant language groups. Now, if you ordered these
via optimality theory, and, say, had a rule that states "verbal phrases
should follow subject/object phrases", and ordered that one first, and
then had a second rule stating, "subjects and objects are identical",
the three top word orders would still be (in order): SOV, SVO and
VSO--even though SVO would be the "odd" one. (A third rule would have
to be "objects should follow subjects", but that would put VOS last,
rather than fourth.)
> Anyway, why all this? It would just posit a way to think of sentenceswhere objects aren't dominated by verbs, but where, rather, verbs
specify relationships between nouns.
> This gave me the idea for a verbless conlang, and I was wondering ifit was similar to All Noun. Here's the idea:
> Slot 1 = NP
> Slot 2 = case particle
> Slot 3 = NP
> Slot 4 = conjunctive particle
> Slot 5 = NP
> Slot 6 = inverse (optional)
> Anyway, so what you'd get is first a noun (let's say, "boy"), then acase particle, which could be accusative (NP1 acts on NP2), locative
(NP1 is stationed at/near NP2), genetive (NP1 is of NP2), etc. Then
you'd have the second noun (let's say, "dog"). Then would come a
conjunctive particle, which would be something like "by means of", or
"under the condition of", or "because of". Then the last NP would be
the verbal NP, which would be some sort of abstract noun (let's say,
"sight"). And then if you wanted to fix a certain NP order, I threw in
an inverse particle, while I was thinking of this. I just realized
it's probably not necessary. Anyway, so what you could end up with is:
> (1) The boy (2) stimulus* (3) the dog (4) by means of (5) sight.
> "The boy sees the dog".
> *I use this for an accusative for experiencer verbs. If I were to putan accusative in there, I'd think of it not as "seeing", but as
> Is this something like how All Noun worked?
> I was just curious. :)
I don't know about All-Noun, but Kélen works sort of like that. Sort of.
For "The boy sees the dog.":
The boy is the experiencer, seeing is the experience, and the gog is the
source of the experience. The relational SE governs experiences, with
particles overtly expressing source and goal (experiencer).
sema jarróñien mo mamóíñ to jakié;
se+3p.sg.anim.goal seeing(N.distr, so it taking place more than once or
over a period of time) goal-marker boy-N.anim.sg
The relational, which is more-or-less a verb is first, though, giving
Kélen language info can be found at:
This post may contain the following:
á (a-acute) é (e-acute) í (i-acute)
ó (o-acute) ú (u-acute) ñ (n-tilde)