Re: The Glyphica Arcana

On Thu, 15 Dec 2005 17:00:16 -0500, Jim Henry <jimhenry1973@...>
wrote:

>On 12/15/05, Jefferson Wilson <jeffwilson63@...> wrote:
>> Jim Henry wrote:
>>> On 12/15/05, Jefferson Wilson <jeffwilson63@...> wrote:
>>>>Jim Henry wrote:
>
>>> Cecil eats noodles.
>>>   Cecil = agent,
>>>   noodles = patient
>>>
>>> Cecil sleeps.
>>>   Cecil = experiencer
I don't think Jim's is what's usually considered the best example of what's
usually considered the Experiencer vs. the Agent; although it turned out to
be worth talking about.

So, I think I should stick this in for Jefferson, just in case;
Experiencer/Stimulus bivalent verbs are often treated differently from
Agent/Patient bivalent verbs.  The difference may be marked by Case, by
Voice or by both.
According to Blake's "Case", there are four sets of
roles, of which each set usually gets handled by the same case in each
particular language, but different languages may handle differently,
although many of them handle them all the same as Agents.
They are:
Agent in Agent/Patient
Perceiver in Perceiver/Stimulus (Sense verbs; see, hear, smell)
Experiencer in Experiencer/Target (Emotion verbs; love, fear)
and I forget the fourth one; it may be Exister or Sitter or Stander or
something -- but don't rely on that guess.

I'm about to time out, or I'd look up the messages about these roles, and
the conlangs that handle them, for this messsage.

>> If the intent of the sentence is to show that Cecil is
>> experiencing sleep, then the name "Cecil" would be marked as an
>> indirect object with "sleep" as a verb.  (At one point I called
>
>_Indirect_ object!  Interesting.  A number of languages
>use the same case for at least some objects
>of transitive verbs and subjects of intransitive verbs,
>but I don't know if there are any natlangs that
>use the indirect object case for subjects
>of intransitive verbs.  Neat.
Yes.  Take a look at Message 134110
(From: tomhchappell <tomhchappell@...>
Date: Wed Sep 28, 2005  8:48 pm
Subject: Re: Thoughts on Tarsyanian verbs
)

It contains the following paragraph:
"
If an Active/Stative or Split-S or Fluid-S language, in which the
nominals of intransitive verbs are case-marked as A or P differently
depending on whether the verb is considered Active or Stative, is
also a Basic Voice language, this makes Klaiman's points easier to
make. He points this out with several Native American languages.
One of them, a Muskogean language called Alabaman, actually has a
three-way split in intransitive monovalent clauses; the nominal can
be ergative, or accusative, or dative. At least one verb, whose
gloss is "to be high up", can take its argument in all three cases.
"It-ERG is-high-up" means roughly "Wow! Look at that thing shoot up
there!"; "It-ACC is-high-up" means roughly "Gosh, is that thing
sitting at the top of that tree, or what?" while "It-DAT is-high-up"
means roughly "Well, /I/ sure can't reach it; can /you/?"
"

I was very interested in this because of wondering why a language with
Split-S (split-intransitive) alignment of monotransitive-to-intransitive,
and "Split-O" alignment of ditransitive-to-monotransitive, wouldn't allow
the dative or dechticaetieative case in intransitive sentences; it turns
out some of them do.

> this 'indirect voice.' That is, a verb and indirect object making
>> a complete sentence.)  However, if the intent of the sentence is
>> to show that Cecil is performing the activity of sleep, "Cecil"
>> would be marked as subject.  "Cecil" might also be marked as a
>> direct object, indicating that "sleep" is being imposed.
>> ("Taking the medication, Cecil sleeps.")
>
>So the "subject" is always some entity taking
>deliberate action?  Maybe it should be called
>the "agent" case instead.
I think Jim is right about this, Jeff.
Look through our recent archives for discussions about "semantic
roles", "thematic roles", and "grammatical relations".

Tom H.C. in MI

Reply