Where  have  Southern  Nilotic  Voiced Consonants  gone?
          -  Origin of  a Falling Tone  in Kipsikiis  -

                                           HIEDA, Osamu
                                             <hieda@osaka-gaigo.ac.jp>
                                             
  This is an attempt to examine the relationship between consonant types and 
tone and to inquire the origin of a falling tone in Kipsikiis.  
  In Kipsikiis stems with a falling tone  (underlying) are classified  into 
two types;

     (1) the one ends with /y/, a diphthong,  or a  long vowel,
     (2) and the other ends  with a  sonorant.

  (1)  The  stems  of  this  type  ended originally in a voiced consonant, and
bore a high tone during the proto-Nilotic period.  The voiced consonants 
changed to a glottal fricative h, and then the h disappeared,  being replaced 
by a falling  tone on the preceding syllable at the stem-final position
  (2)  The Kipsikiis has the tonal processes  through which surface 
representations  are derived from underlying tonals tructures.  One of these 
processes  was that a high tone on a final syllable ending in a sonorant was
1owered to a falling tone (FF).  A falling tone (surface) caused by FF was 
phonologized as an underlying falling tone;  i.e. a falling tone to which an 
underlying high tone was lowered by FF, was reinterpreted as an underlying 
falling tone in monosyllabic stems by speakers, in order to distinguish them 
from the stems which became monosyllabic by the loss of segments.