Re: THEORY: two questions

On Mon, 27 Mar 2000, Matt Pearson wrote:

> Depends on who you ask.  My definition:  A head-marking
> language is one which keeps track of arguments (who did
> what to whom) primarily by means of agreement on the verb,
> while a dependent-marking language keeps track of arguments
> primarily by means of case-marking on the noun phrases.
> That is, "head" = the verb, while "dependents" = the noun
> phrases.
>
It was my impression that the terms also applied to grammatical
categories besides just verbs -- that is, prepositional phrases,
possessive constructions, and noun phrases can be classified as
head-marking or dependent-marking as well.  Welsh's inflecting
prepositions would be a good example of head-marking in a prepositional
phrase:

        arna  i;  arno   fo;   arni   hi
        on-1s me  on-3sm him;  on-3sf her

Here the head of the prepositional phrase (the preposition) is being
marked, instead of the dependent -- that is, the head carries the
information about the relationship between the head and the dependent.
This would contrast with something like, oh, say, Latin, where the head
(the preposition) remains invariable and the dependent (the noun) gets
the morphology:

        in silvam;       in silva:
        in forest-abl    in forest-acc
        into the forest  in the forest

Then there are the languages that are head-marking in possessive
constructions, the classic example of which is Hungarian:

        az  ember haz-a
        the man   house-3s
            (dep) (head)

This contrasts with English, which has the morphology on the dependent
instead of the head:

        the man's house
            (dep) (head)

Finally, there's head-marking between nouns and their adjectives, which
totally blew me away when I saw it in class.  Of course, languages that
are mark the dependent adjectives in NPs are a dime a dozen, i.e.
French:

        l'homme heureux;     la  femme heureuse;   les femmes heureuses
        the man happy-s.masc the woman happy-s.fem the women happy-pl.fem

But what blew me away was that apparently, in Farsi, it goes the other
way -- the head noun is marked when there are dependent adjectives (this
example is actually taken from another article by Nichols, heh):

        kuh-e        boland
        mountain-DEP high
        (head)       (dep)

This <-e> suffix is apparently only used when the noun has an adjective.
And apparently this structure is pretty rare cross-linguistically, but
it's cool beyond imagination, and I'm convinced at this point that it
deserves a place in Doraja.  For example:

        dae enin geda
        I   like cheese
        I like cheese.

        dae enin geda-i kyla
        I   like cheese-DEP blue
        I like blue cheese.

I'd also probably have it for nouns with relative clauses:

        dae enin geda-i     sao      sen te fes
        I   like cheese-DEP that-nom be  on table
        I like the cheese that's on the table.

Or something like that.  (Comments on this?  Can anyone off-hand think
of other natlangs/conlangs that work this way?)

Later,
Adam