Hiksilipsi Consonant Phonology

Hey, y'all.  I've been working on this for a while, and I decided to come
out with what I have of the phonology for now.  Any comments and/or
suggestions are welcome.

Hiksilipsi contains three grades of phonemes: a radical grade, a y-grade,
and a w-grade.  The y- and w- grade phonemes are formed by adding a y [j]
or w [w] glide to the radical phonemes, but there are several mutations
that occur when these grades are formed.  To begin, the radical phonemes
are (written with the orthographic symbols, with the IPA next to it where
different.  Thanks to YHL for giving me the idea to spell [?] as {q}):

        p        t         k        q [?]
        f [P]   s                   h
        ps                 ks
        m        n        ñ [N]
        mp                ñk [Nk]

The nasals and nasal occlusives do not participate in the mutation
scheme, so I'll leave them out of the next charts.

When combined with the y-glide, the phonemes take the following forms.
Throughout this chart {y} represents [j]:
        py        ty [tç]          qy [?j]
        fy [Pj]  sy                hy [ç]
        phy [pç]

Thus, the most salient changes are /ty/ > [tç], /hy/ > [ç], and /psy/ >
[pç]

The w-grade phonemes are:
                tw        kw        qw [?w]
        hw   sw                     hw [w_0]
                           khw [kw_0]

The only really important changes are /fw/ > [w_0], /hw/ > [w_0], and
/ksw/ > [kw_0].

You'll notice that there are no velars with y-glides or labials with
w-glides.  This is because there is an active dissimilation in
Hiksilipsi: when /y/ is preceded by /k ks ñ ñk/ it becomes /w/ and
affects the preceding sound as /w/.  Likewise, when /w/ is preceded by /p
ps m mp/ it becomes /y/ and affects the preceding consonant as /y/.  The
only exception to this rule is /f/.  Underlying /fw/ does exist, because
very early in the language there was a rule that changed [P] to [h] when
followed by a rounded sound.  Thus, /fw/ merges with /hw/, as shown
above.

There are a couple of odd things about this system.  Everything's pretty
regular underlyingly, but the surface forms are really weird.  There is
no plain [w], but there is a [w_0].  There are the clusters [?j] and
[?w], which I've never seen in a real language.  Then /psy/ and /ksw/
mutate to assimilate the /s/, but plain /sy/ and /sw/ are unchanged.

Do these things seem reasonable to everyone?  Or is it hopelessly weird?

Jesse S. Bangs     Pelíran
jaspax@ juno.com
"There is enough light for those that desire only to see, and enough
darkness for those of a contrary disposition."  --Blaise Pascal

Reply