Process-morpheme stems

[Tuggy, D.]

     I received a number of interesting and useful replies to my question
     about possible process-morpheme stems (for which, Thanks).  There were
     a couple of good examples of zero stems, but no process-morpheme stems
     yet. Several respondents questioned the coherence of the idea.  For
     instance Nick Kibre: "I'm having trouble thinking of what such a thing
     would be; what would a process morpheme root's process be carried out
     on?" Or David Stampe: "If a morphological process normally operates on
     a base to produce a derivative or inflection, since a root or stem is
     such a base, what could it operate on?"

     The answer is on whatever is next to it, of course.  Usually this
     would be an affix, though I suppose it could conceivably be a
     compounded stem or even an adjacent word.

     The sort of thing I have in mind is this: we recognize a zero stem
     when the affixes that would usually attach to a stem seem to attach to
     nothing; either the rightmost prefix attaches to the leftmost suffix
     (and vice versa), or there are only prefixes or only suffixes there.
     (Of course there will typically be meaning components which logically
     pertain to a stem, as well.) It is at least conceivable that a
     language might have two such zero stems, differentiated phonologically
     in that one of them triggers one tone pattern in the affixes and the
     other another pattern, or one palatalizes and the other doesn't, etc.
     (It wouldn't really be necessary for there to be two such stems in
     contrast, as long as the change in the affixes could be recognized as
     such.) Any process morpheme can be thought of as a zero that betrays
     its presence by producing some change in the neighboring morphemes.
     Such change-triggering is of course not limited to zeroes, nor is it
     limited to affixes--lots of stems do it. I just don't know of any zero
     stems that do it, only zero affixes.

     The existence of process-morpheme stems would tend to imply a
     less-than-absolute division between stems and affixes, which would be
     no big problem to me (in fact positing an absolute division would be
     quite problematical on other grounds.) But of course the really
     interesting question is the empirical one: are there any out there?

     Hope this clarifies the question a bit.

     David Tuggy