Re: Eliding repeated morphemes: synthesis vs analysis

On 27 Dec 2004, at 5.20 pm, Thomas R. Wier wrote:

> > We had a discussion about this some months back about the
> > technical distinction between case affixes and clitic postpositions.
>
> You don't perchance have the thread title to you? Would aid searching.
Search for "a case-free language"; we discussed under that thread
in early October.

> > The most widely accepted criteria are those set forth by Zwicky and
> > Pullum:
> >
> > (1) Clitics have “freedom of movement”, affixes do not.
> >    (where 'movement' need not necessarily imply Movement
> >     with a capital-M.)
> > (2) Clitics can attach to material already containing clitics; affixes
> >     (since they are morphological entities) are pre-syntactic and
> >     cannot attach to material containing clitics.
>
> 'If it attaches to a clitic, it is a clitic'?
That is, things which are felt to be clitics by other criteria.
There are few if any known examples of something which is clearly
an affix attaching itself to something which is clearly a clitic.
One possible exception would be the sequence in "I'd'n'a Xed"
(= I would not have Xed), where by the Zwicky and Pullum criteria
the /n/ negator is attached to the clitic verb stem /d/, but this
certainly isn't a clear case.

> > (3) Clitics have freedom of host selection, affixes have no freedom of
> >     stem selection.
> > (4) Clitic-host combinations may not have idiosyncratic meanings;
> >     stem-affix combinations may.
> > (5) Clitics may neither trigger nor undergo morphophonological or
> >     suppletive alternations, affixes may.
>
> I thought /s~z~@z/ was a morphophonolgical alternation? Am I wrong,
> or are these just guidelines?
Right! This is why I stated (in a segment of my post that you clipped)
that clitics are a heterogeneous class:  they may by five criteria
count approximate the clitic ideal, but by one criterion the affix
ideal.

> (6) Clitic-host combinations may not have arbitrary gaps;
>     stem-affix combinations may.
>
> By which I assume we mean "you can attach +im to any word except
> prepositions" would invalidate it? (assuming +im is a clitic &
> prepositions could come in a position where +im may attach).
No, it means more that clitics will attach to any member of some
syntactic class, and will not e.g. attach to some lexical items
but not others.

 =========================================================================
Thomas Wier	       "I find it useful to meet my subjects personally,
Dept. of Linguistics    because our secret police don't get it right
University of Chicago   half the time." -- octogenarian Sheikh Zayed of
1010 E. 59th Street     Abu Dhabi, to a French reporter.
Chicago, IL 60637