Re: Messy orthography (Re: Sound change rules for erosion)

John Cowan wrote:
> Or as in the case of Maori verb inflection, where essentially all
> the verbs in the language except recent borrowings are irregular!
> Simplification onto the regular endings is in progress.
>
It isn't as bad as it sounds.  All the Polynesian langs (and Fijian too)
intercalate a consonant when they add the "transitive" and/or "passive"
suffix-- but the number of consonants is quite limited-- by and large, the
reflexes of the Proto Fi-PN finals, which were *p t k m n N r l s.  Due to
mergers (*r > l in PN) and rarity (*p) not all possibilities occur, and as
you say, there has been some analogical reshuffling and a tendency toward of
regularization , such that these intercalated C may not necessarily reflect
the historic final anymore.

Very likely, there are Melanesian languages (both MN and PN are in the same
Oceanic subgroup) that do the same. A handful of MN languages actually have
preserved some of the finals, usually by adding a support-vowel.

Various langs. within Indonesia which have also reduced the original canonic
CVCVC form to CVCV have similar irregular suffixing procedures. It suggests,
of course, that the finals _may_ still be present underlyingly, and that
their loss is _relatively_ recent.  But most speakers, I think, would claim
that the consonant belongs to the suffix, not to the root.

Buginese is a rather extreme example-- all final stops and liquids have
reduced to /?/, but when you add the verbal suffixes, the /?/ changes
to -r-, -s-, or -k-.  Most forms consistently choose just one; quite a few
can have two, a handful have all 3, sometimes with slight changes in
meaning.  The situation has clearly changed since Matthes' 1853 dictionary--
my younger informants frequently gave forms that didn't match the
dictionary's.

Reply