quirky case (was: Re: New Language Sketch)

Tom Wier wrote:

>Speaking of case marking, does anyone use quirky case like
>Icelandic does?  Some Icelandic verbs require really odd cases
>not just for objects (anyone who's learned Latin or Greek is
>used to those), but for subjects too.  For example,
>
>(1) Hana       vantar mat.
>     she:ACC lacks   food:ACC
>     "She lacks food."
>
>(2) Barninu           batna=F0i             veikin
>     the.child:DAT  recovered.from the.disease:NOM
>     "The child recovered from the disease."
>
>(3) Verkjanna         g=E6tir            ekki
>      the.pains:GEN  is.noticeable not
>      "The pains aren't noticeable."
>
>In all of these examples, the first word is the subject -- and yet
>each verb requires a certain case marking as a lexically determined
>feature of the verb.  Fun!
Subject case-marking in Tokana is super-quirky.  Subject can
take ergative, absolutive, dative, ablative, or instrumental
case, depending on the verb.  There are certain patterns, though:

(1)  Verbs denoting volitional activities generally take ergative
subjects (although there are numerous exceptions).  Note, though,
that the ergative case can only be used for definite animate
subjects.  If the subject is inanimate, it's marked with instrumental
case.  If the subject is animate indefinite, it's marked with ablative
case.

(2)  Motion verbs take ergative subjects if their meaning
emphasises manner of motion ("walk", "run", "swim", "fly")
and absolutive case if their meaning emphasises trajectory
of motion ("come/go", "arrive", "leave", "enter", "go
through").

(3)  Stative verbs typically take absolutive subjects.

(4)  Psych verbs - i.e., verbs denoting emotional or
epistemological states - typically take dative subjects, although
there are some important exceptions ("love" takes an ergative
subject, for example).

(5)  Auxiliaries are idiosyncratic:  "can" and "must" take
ergative/ablative subjects, "become" takes an absolutive
subject, and "intend to" and "want to" take a dative subject.