social benefactive/antisocial malefactive case?

[jess tauber]

Hi folks. I have questions about case. Working with data from Yahgan (isolate, Tierra del Fuego) I'm finding that various case markings have rather specific uses in certain circumstances. I'm not used to case-marked languages, as almost all the languages I have experience with are either unmarked or primarily head-marking.

In Yahgan there is an accusative form -ima which has a very antisocial, malefactive usage when the noun phrase is singular and human. Plurals with this mark don't have to be human, nor is the sense necessarily malefactive. Yet there is another form -nchi- used only when there is a generally positive socially defined and specifiable human relationship (unequal, such as between master/servant, king/subject, teacher/student, etc.) and no party is completely free to do as he/she wishes except within the normal bounds of the social "contract" implied by the noun dyad.

Obviously this is all tied up with definiteness, person, number, animacy/control in general.

So, is this NORMAL? Or is this rather unusual in case systems? One sees such a malefactive/benefactive split in the voice system of Salishan languages (as suffixes on the verbal head), and I claim that Yahgan and Salishan are in fact more closely related than languages in between (is that a wink from Sapir??). Is this normal for head-marking languages as well?

Thanks for any tips,

Jess Tauber