Re: Branching typologies

At 7:07 PM -0400 09/27/01, David Peterson wrote:
>In a message dated 9/27/01 11:46:10 AM, Dirk_Elzinga@BYU.EDU writes:
>
><< The resulting
>structure is somewhat flat, rather than left- or right-branching.
>Since the verb is normally in first position, I suppose you could
>call it verb-initial. >>
>
>     How are the morphemes arranged?
I have never been convinced of the generalizability of the order of
morphemes within a word (prosodic or grammatical) to word order in
general. Hence my statement regarding Tepa clause structure as being
flat.

That said, inflectional morphemes are arranged in the following order
(I haven't worked out derivation; I have some ideas that I'm not
ready to commit to or share yet):

      {determiner,mood}=person-transitivity-STEM-(tense/aspect)

All Tepa predicates are introduced by a modal clitic (in the case of
matrix verbs) or a determiner (in the case of adjoined predicates).

The person prefixes conflate the notions "subject" and "object" into
single unanalyzable forms; their interpretation will depend on the
presence (and type) of transitive prefix. There is thus no basis for
determining the order of subject and object arguments within a
clause. I'm not sure yet if there is a preferred order for adjoined
predicates which are coindexed to arguments. I'm guessing that there
isn't since the adjuncts are optional (Tepa is a pronominal argument
language).

A lot of the most interesting morphological action takes place in the
STEM. Prosodic strategies are employed to mark number and
tense/aspect distinctions; the suffixed t/a is present only under
prosodically defined conditions. So it is hard to say what order
these "morphemes" come in.

Dirk
--