Re: THEORY: "Quirky" Case -- "Quirky" Subjects and "Quirky" Objects

Hello, David, thanks for writing.

--- In conlang@yahoogroups.com, "David J. Peterson" <dedalvs@G...>
wrote:
> Tom wrote:
> <<
> David, you said Latin had case-governing verbs for each case except
> Vocative, such that the verb in question requires its subject to be
> in that case.
> Is there a Transitive case-governing verb for each Subject case?
>  >>
>
> I'm pretty sure the answer is "yes", but I can't find my paper
listing
> what exactly the cases of Latin did (and the same goes for objects).
>
> Tom continues:
> <<
> I forgot to add my "signature" question about Ditransitives;
>
> 3a) In ditransitive sentences, how many, and which, grammatical
cases
> can be assigned to each of the core arguments --
> grammatical/syntactic subject, grammatical/syntactic direct object,
> grammatical/syntactic indirect object?
> 3b) For each of the core arguments, which language holds the record
> for most grammatical cases assignable to that grammatical/syntactic
> role? And what is that record number?
>  >>
>
> First, I have to say that these questions strike me as rather
bizarre.
> Why would it matter if a given language has the largest number of
> cases that can be assigned to role X?  Or the fewest?  Or an average
> number?  Additionally, why does it matter which cases?  If you can
> come up with a story for it, then that pretty much licenses it in a
> conlang.
Sorry about that.
Read my answer to Markus Miekk-Oja to see what I know now that would
have kept me from asking such "bizarre" questions.
I have been reading a couple of articles by Paul Kiparsky of Stanford
University and by Halldor Sigurdsson of the University of Lund.
http://kybele.psych.cornell.edu/LU/Kiparsky-Cornell.pdf
http://mercury.hau.ac.kr/kggc/Monthly/ReadingMaterial/Case.pdf
They helped me understand.

>
> Anyway, one paper I recommend reading is one I referred to in
> my ergativity reference, but which, heretofore, was apparently
> improperly linked to, and, even more embarrassingly, was
misattributed
> (don't know *where* I got "Stepher" S. Dryer).  It's a small,
readable
> paper on clause types that deals with ergativity and  ditransitivity
> by Matthew Dryer.  It can be downloaded from my site here (because
> I can't find it on the web anymore):
>
> http://dedalvs.free.fr/misc/clausetypes.pdf
>
> Plus, if anyone can find a place to download it on the web, let me
know.
>
Dryer was apparently the person who first asked and started to answer
questions about the alignment of ditransitive with monotransitive.
Haspelmath is the author of what seems to be the latest authority.
I want to read Dryer's work when I get the chance; that won't be
until later today at the earliest.
Thanks for the URL.

> Tom:
> <<
> Are David and Markus saying that in "Quirky Case" languages, the
> grammatical/morphological case assigned to any core-argument
nominal,
> is wholly semantically determined (by the lexical verb among other
> information), not ever just a function of the grammatical/syntactic
> role it has as a core-argument?
>  >>
>
> Well, to what extent is *anything* purely syntactic?  If there's a
> rule that subjects take the nominative but there are a handful of
> counterexamples that take the dative, then is it the case that a
> syntactic subject takes the nominative case with some lexical
> counterexamples, or is it the case that verb lexically assign case
> to arguments which occupy a certain syntactic position?  Anyway,
> I suspect that much that's considered purely syntactic probably
> has semantic/morphological aspects to it, as well--or even
> phonological (e.g., "Screw in the lightbulb" sounds fine, but "Screw
> in it" for the same intended meaning sounds a bit bizarre, and
> the reason is probably just phonological).
Sigurdsson's paper helped me understand this.

>
> Anyway, what are you looking to do?  I'm curious.
>
When I first read your question, I guess I didn't know yet.
But what I was trying to do was come up with a case system for my
first conlang.
As a result of this discussion I have decided on a four-case system
for the "structural" or "argument licensing" cases; I will decide
later whether or not to have more cases for the "inherent" cases.
My system will be tripartite from the PoV of bivalent-to-monovalent
or monotransitive-to-intransitive alignment; it will be
indirective/directive from the standpoint of ditransitive-to-
monotransitive alignment.
I will call the major cases Absolutive, Ergative, Accusative, and
Dative.
Every clause that has a core argument must have exactly one core
argument that is either Absolutive or Ergative.
Every clause that has a core argument must have exactly one core
argument that is either Absolutive or Accusative.
For any monovalent clause, its only argument must be Abolutive.
For a transitive clause, its Agent will be Ergative, and its Patient
will be Accusative.
For ditransitive clauses, the Recipient will be Dative and the Theme
will be Accusative.

I have decided to have a set of verbs that license two arguments, of
which one must be Absolutive and one must be Dative; I am going to
call these Bivalent Intransitive verbs.
Their English glosses will prototypically have to include a reflexive
and a preposition.
Example; I'll have a verb with the gloss "concern oneself with".

Thank you for writing.

Tom H.C. in MI