Coronality, velarity and why they are special

This  paper addresses the problem  of  the representation of
places of  articulation  of consonants in  the  framework of
Government  Phonology.  The  emphasis will be  on coronality
and problems GP encounters when it has to give an account of
the  asymmetries  in the  behaviour  of  coronals.   This is
hypothesised to be  a consequence of  their unmarked nature,
their lacking  a specification   of place  of  articulation.
This hypothesis is supported by a number of phonological and
non-phonological phenomena many  of  which are presented  in
Paradis &  Prunet  (1991).   Velarity will  also  be briefly
discussed because this is the consonantal place the GP model
suggests  to be most unmarked and  I conclude that there are
serious  problems  with consonantal  representations in  the
model.

Section  1   introduces the  basics   of  GP crucial to  the
understanding   of the later  sections.    In 2 evidence  is
brought up  to  prove that  coronality  is special among the
other places of articulation: there is a coronal/non-coronal
dichotomy.   The problems GP  faces  in expressing this fact
are discussed in 3, with an investigation of what happens if
the coronality element  is discarded in 3.1.  The speciality
GP tacitly attributes to velarity is pointed out in 3.2, and
some conclusions are drawn in 4.