Re: A mixed ergative question...
|From:||Kala Tunu <kalatunu@...>|
|Date:||Saturday, April 13, 2002, 7:24|
Your system is quite nice. I think your NOM, ACC and DAT actors are very much
like Rick Morneau's Agent, Patient and Focus actors:
(He:NOM - kicked) (became wounded - him:ACC) = he kicked him, injuring him.
= He-AGENT kicks [and/in order to] wounded he-PATIENT
(He:NOM) (died - him:ACC) = he killed him.
= He-AGENT [acts and] dies he-PATIENT
(knows - he:NOM - french:DAT) = he knows french.
= Knows he-PATIENT [with/about] french-FOCUS
As you point it, your examples show that like RM's Patient your own ACC actor
may be the subject of a passive state ("to die"), dynamic state ("to learn"),
dynamic uncontrolled state ("to sneeze"), etc., and like Rick Morneau does too
you seem to consider your ACC actor (RM's Patient) as the default actor of
I am not going to suggest you anything--and i'm sure you already thought about
all this anyway--but still i was thinking that a language based on Causation
should tell Causation from Goal. Maybe this could be achieved by tagging
factitive/lassive Agent (FNOM) differently from the plain transitive Agent
(TNOM). In other words, your NOM subject would be either factitive or plain
transitive, just because that disctinction is tightly linked to the distinction
between Causation and Goal. Let me develop:
(i) Sometime the action of the Agent is not specified:
"AGENT [acts somehow and] PATIENT dies"
This is called a prospective aspect because you qualify the verb according to
the resultive state of the patient, not the causative action of the agent: the
verb is Factitive.
(ii) Sometime the result of the Agent's action is not specified:
"AGENT kicks PATIENT [but who knows what entails?]"
That is called a retrospective aspect because you qualify the verb according to
the causative action of the agent, not the state of the patient. The verb is a
plain Transitive (my fave french semanticians call this one "Exotropic" because
it "turns outside" the subject "towards" the object--but there are also
"Endotropic" and "Atropic" ones as i have been rabitting it on for a while now
just in case this would subliminally get across some conlanging souls... :-))
(iii) Both Factitive/Lassive and Transitive verbs can be Transitive or
Ditransitive--which is also why i think the current "transitive vs.
intransitive" label is shabby--:
"The master teaches sciences to the student.": This is a Ditransitive Transitive
with result unknown: does the student actually learn sciences or rather indulge
in conlanging during classes?
(iv) "The master makes/has/lets the student learn sciences.": That is a
Ditransitive Factitive/Lassive focusing on the student's dynamic state--although
everything depends on the student's will...:-)
(v) So certainly i lost you there ;-) but my purpose was to show that your lang
could maybe tag Causation and Goal differently. Although i guess you can express
the difference with REALIS ("...resulting in...") and IRREALIS ("...aiming
at...") tags as well.
Actually, i find it odd that most auxlangs take for granted that the "obvious"
or "natural" subject of some verbs is the Agent and of some other verbs is the
Patient or the Focus. For instance in japanese the "obvious" subject of "to
break" is the Patient broken or breaking, while in french it is the Agent
breaking the stuff. The "obvious" subject of "to transfer" is the transferring
Patient in japanese and the Agent transferring in french. While english proves
it is the best possible auxlang by managing to just not make a decision ;-)
(HUMOUR: auxlangers, no beating up please!)
oops! this is a looooong post and now i'm late!!! sorry for rambling on
Garrett Jones <alkaline@...> wrote:
>>>I am also experimenting with ergative stuff for my own language Minyeva
(previously Malat). The system i have at the moment is nominative in case
marking but ergative in constituent order, which i understand doesn't occur
in any natural languages. It happens the other way around (ergative case
marking, accusative word order) but not the other way.
Minyeva is being designed around a regularized causation system, since i
noticed that causation is generally done unpredictably across languages (but
not necessarily inside one language). The language would have state verbs
(both static and dynamic) and intransitive action verbs pre-subject, but
transitive action verbs post-subject. Sentences are a string of "subjects"
with their verbs attached to them, in a type of causation sequence. It would
look like this:
(state - agent - action) - (new state - patient - action) (new state -
patient - action) and so on.