Re: typology of V-initial lgs

Kristian Jensen wrote:

>This is what I have pictured for Boreanesian. The Boreanesians migrated
>from Japan before sea levels rose over 10K years ago. Japanese, Korean,
>and Ainu are all verb-final. So Boreanesian was perhaps a verb-final
>language. But due to an areal influence from the surrounding Austronesian
>languages, Boreanesian is now a verb-initial language with a post-
>positional holdover from ancient times.
Hmm.  Perhaps I can appeal to diffusion to explain some unusual word
order properties of Tokana.  Tokana is verb-initial (VOS being the unmarked
order in embedded clauses), but it does have some 'deep' morphological features
which suggest that it might have originally been verb-final, to wit:

-- Tokana is predominantly suffixing (very few prefixes) - tense/aspect,
        noun case, etc., are all indicated by suffixes
-- Lexicalised compounds are head-final
-- Negation takes the form of a suffix on the verb
-- Clausal complementation is marked by suffixes on the embedded verb

These last two features are, as far as I know, virtually unattested in
verb-initial natlangs.  Perhaps Tokana was indeed originally verb-final,
and became verb-initial as a result of areal influence.  (I don't know
anything about Tokana's neighbour languages, except that they are
completely unrelated to Tokana, so that might lend some credence to
the diffusion story...)

Coincidentally, linguistic areas and morpho-syntactic diffusion are the
topics of my lecture today in the Historical Linguistics class I'm teaching!

Matt.