On Thu, Aug 7, 2008 at 11:20 AM, David McCann <[log in to unmask]> wrote:
> On Wed, 2008-08-06 at 09:35 -0400, Jim Henry wrote:

>> Is there a standard
>> term in linguistics for this theta role or case?  Are there natlangs or

> I couldn't remember any examples, and a quick check of Palmer's
> "Grammatical roles and relations" doesn't show any sentences of this
> type, but I should think that it would be covered by the distinction
> between affected and unaffected patients.

"Unaffected patient" seems at first like a contradiction in terms, as
far as I understand the term "patient".  But:

> In Ga'dang (Austronesian), the verbal marking shows if the patient (when
> the topic) is altered (e.g. broken) or not (e.g. tied up).

That makes sense, and I guess affected vs unaffected patient
is not a bad way to terminologize it.  Still, gzb would conceptualize
tying a knot in a string as a patient relation (or perhaps
object-of-result, if the object noun were knot rather than string)
even though the string isn't as drastically affected by tying
as it would be by cutting or burning.  If I were starting over,
I might have two or more patient cases for different degrees
of affectedness, but this part of gzb is too stable for that
to change now.

> It seems too esoteric to ever get its own case or verbal marking (and so
> to need a term in linguistics), but the object of a performance could
> obviously be treated as a dative or partitive, or put into an oblique
> case as in gzb. The performer would seem to be nominative or ergative:
> after all, they are *doing* something, unlike one who experiences a
> sight or liking.

Yes, the performer of all these verbs would get agent-case marking
in gzb.  The (tentatively named) performative case isn't necessarily
oblique in gzb, though; I'm not sure gzb has any oblique cases
as distinct from non-oblique cases.  All cases except topic,
agent and experiencer are obligatorily marked by postpositions;
the patient case doesn't get special treatment that marks
it out from other theta roles, like the accusative case in
IE languages.  I think the only way you could say that the
patient case postposition is special is that it occurs more often in the
corpus than the other postpositions; syntactically it's treated like all
the other object case postpositions.

The most common case and spacetime postpositions in the corpus are,

 2.7230%	218	hxy-i			patient
 2.4731%	198	miq-i		topic
 2.2733%	182	i			at
 1.4864%	119	kax-i		object of attention
 1.4614%	117	tu-i			agent
 1.2366%	99	o			to
 1.0867%	87	jax-o		becoming
 0.8619%	69	sqi			after
 0.8119%	65	nxiqn-i		comment
 0.5121%	41	im			part of
 0.4871%	39	nxaxw-o		addressee of communication
 0.4372%	35	daxm-rq		by (authorship)
 0.3997%	32	jax-rq		ceasing-to-be
 0.3747%	30	kujm-o		for, in order to
 0.3622%	29	liqw-i		related to
 0.3123%	25	kriq-o		object of result
 0.2873%	23	sxu-i		of (quality)
 0.2748%	22	il			through
 0.2623%	21	jqaxr-i		experiencer
 0.2498%	20	syj-i			with, using
 0.2498%	20	jax-i			in such a state

I'm not sure how good a corpus this is re:
representativity, since it has a mix of archaic and current
text and I haven't made any effort to balance the
relative word-count of different genres as in the
Brown Corpus.  About half of it is randomly selected
journal entries from 2002-2008, and most of the other
half is translation exercises and relay texts.


> Of course, some people want a word for *everything*: generally in wikis
> rather than in books, though.

I need some way to refer to this distinctly in the gzb grammar, and I'd like
to use a standard term if I can.  I can make up my own terms like
the Lojbanists, of course, if I have to, and it looks like I'll have to in
this case (pun unintended, but shamelessly left in place once noticed).


On Thu, Aug 7, 2008 at 11:31 AM, Yahya Abdal-Aziz <[log in to unmask]> wrote:
> Jim, how important to your distinction is the degree of creativity involved?
>
> From the perspective of modern (and perhaps, post-modern?) performers,
> each performance (singing, playing, reading, presenting) of an established
> work is an original creative act, of no less significance than the creative act
<snip>
.........
> sufficiently tax my resources!  I guess my point is this: that although I can
> see the distinction you're making, in theory, it doesn't seem to be of any great
> practical significance.

I think in gzb the degree of creativity is not what matters, but the
transitoriness of the creative act's result.  I would use {krĭ-o} to
mark objects-of-result of creative acts that stick around in some
form or other, even for a relatively short time after the action
of the verb is complete, and {ĉul-i} for creative objects-of-result
that exist only while the action of the verb is being done.

As for the practical significance, consider that gzb, though not
verbless like Kelen, is (what I might call) verb-drop; if the postpositions
in a sentence are explicit enough about the relations of the
various entities denoted by the nouns, then no verb is needed.
So for instance

    ŝrun-twâ  	 ĉul-i.
    music-saying 	performance-at
    I sing a song.

    ŝrun-twâ 	krĭ-o.
    music-saying 	create-to
    I compose a song.

Here the explicitness of the two postpositions makes it unnecessary
to supply verbs.  (The first-person agent is also left out by the default
subject rule.)

-- 
Jim Henry
http://www.pobox.com/~jimhenry/conlang/fluency-survey.html
Conlang fluency survey -- there's still time to participate before
I analyze the results and write the article