<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html>
<blockquote TYPE=CITE>
<pre>Matthew S Dryer wrote:
I am interested in finding examples of what I call "floating clitics",
clitics which must attach phonologically to some other word, but where the
host phrase or word cannot be defined in terms of some specific position
in the clause (like second position or final position) or in terms of some
specific host (like subject or verb). My impression is that this sort of
clitic is not uncommon, but does not seemed to be discussed much in the
literature on clitics that I have looked at.</pre>
</blockquote>
Some East Caucasian languages seem to correspond to a mixed type. For instance,
in Udi (South East Caucasian - Lezgian) personal clitics that generally
show an S=A (accusative) behavior are floating 'subject/pivot' indices
AND indicate topicalization, cf.:
<p>(1) bez wich-en ash b-esa-na
<br> my brother-ERG work:ABS
do-PRES-3SG:S=A
<br> 'My brother is WORKING'
<p>(2) bez wich-en ash-ne b-esa
<br> my brother-ERG work-3SG:S=A
do-PRES
<br> 'My brother does a WORK'
<p>(3) bez wich-en ich ash-n-ux k'uax-ne b-esa
<br> my brother-ERG REFL work-SA-DAT2
at=home:3SG:S=A do-PRES
<br> 'My brother does his work
AT HOME'
<p>(4) bez wichi t'eghi-ne ar-e
<br> my brother:ABS today-3SG:S=A
come:PAST-PERF
<br> 'My brother has come TODAY.'
<p>(5) bez wichi-ne ar-e
<br> my brother:ABS-3SG:S=A come:PAST-PERF
<br> 'My BROTHER has come.'
<p>Note that endoclitization (EC) may apply, cf.:
<p>(6) bez wich-en sa khäshil u-ne-k-i
<br> my brother-ERG one pudding:ABS
eat-3SG:S=A(EC)-AOR
<br> 'My brother ATE a pudding.'
<p>There are certain restrictions to the floating of clitics, such as the
negation clitic (!) _te_ which must be followed by the personal clitic
as in:
<p>(7) bez wich-en me ash-n-ux te-ne b-esa
<br> my brother-ERG PROX work-SA-DAT2
NEG-3SG:S=A do-PRES
<br> 'My brother does NOT WORK.'
<p>(8) bez wich-en me ash-n-ux b-esa-te-ne
<br> my brother-ERG PROX work-SA-DAT2
do-PRES-NEG-3SG:S=A
<br> 'My brother DOES not work.'
<p>Note that NEG + PERSONAL CLITIC create some kind of 'clitic cluster'
that itself obeys to the restrictions of clitization, e.g. clitics must
follow the verb in case it is marked by the optative or the 'future1',
cf.
<p>(9) bez wich-en me ash-n-ux b-al-le [-le < -ne]
<br> my brother-ERG PROX work-SA-DAT2
do-FUT1-3SG:S=A
<br> 'My brother WILL DO this
work.'
<p>(9') *bez wichen me ashnux-ne bal
<br> *bez wichenne me ashnux
bal etc.
<p>(10) bez wich-en me ash-n-ux b-al-te-ne
<br> my brother-ERG PROX work-SA-DAT2
do-FUT1-NEG-3SG:S=A
<br> 'My brother WIL NOT DO this
work.'
<p>(10') *bez wichen me ashnux tene bal
<br>
*bez wichen tene me ashnux bal
etc.
<p>The reasons for these restrictions are not yet fully understood. It
seems that both functional aspects (standard topicalization of modality)
and historical aspects (FUT1 and OPTATIV stemming from old inflected infinitives)
apply.- In some other East Caucasian languages (e.g. Tsakhur, Rutul) the
Udi system is paralled by the clitization of class markers.
<p>Best wishes,
<p>Wolfgang
<p>*****************************
<br>Prof. Dr. Wolfgang Schulze
<br>Institut fuer Allgemeine und Indogermanische Sprachwissenschaft
<br>Ludwig-Maximilians-Universitaet München
<br>Geschwister-Scholl-Platz 1
<br>D-80539 München
<br>Tel.: +89-21805343 / Fax: +89-21805345
<br>Email: W.Schulze@lrz.uni-muenchen.de
<br><A HREF="http://www.lrz-muenchen.de/~wschulze/">http://www.lrz-muenchen.de/~wschulze/</A>
<br>*****************************
<br> </html>