syncretism w/o paradigms

[Martha McGinnis]

Hi John,

>It has also seemed to me that impoverishment is naturally viewed as an
>aspect of the syntactic structure -> morpheme structure mapping.
>Under this view, impoverishment happens before the lexicon comes
>into play in filling in the nodes of the morpheme structure.
>
>Are there any knock-down arguments against this view?

I don't think so, but there are arguments that things are a bit more
complicated. Jonathan Bobaljik and Andrew Nevins have both argued in
recent work that impoverishment can be 'interleaved' with vocabulary
insertion into syntactic nodes.  That is, a syntactic feature can
condition the insertion of one vocabulary item, then be deleted
before another vocabulary item is inserted.  This hypothesis makes
interesting predictions about the order of vocabulary insertion.  For
example, if vocabulary insertion proceeds from the inside out (i.e.
root outwards), the prediction is that items that reflect the
pre-impoverishment feature complex will be closer to the root than
items that reflect the post-impoverishment feature complex.

Jonathan's paper is called "The ins and outs of contextual
allomorphy", and the relevant work of Andrew's that I saw was a
conference handout discussing (among other things) person agreement
in Mam (perhaps WECOL 2002?  -- he can give you more useful
information).

Cheers,
Martha
--
mcginnis at ucalgary.ca